Audit recent Agent memory research

data/research/memory-classified.json

@@ -0,0 +1,676 @@
+{
+  "generated_at": "2026-07-10",
+  "model": "ChatGPT-5.6:Sol",
+  "source": "data/research/memory-corpus.json",
+  "total": 293,
+  "completed": 12,
+  "papers": [
+    {
+      "arxiv_id": "2509.02444",
+      "title": "AppCopilot: Toward General, Accurate, Long-Horizon, and Efficient Mobile Agent",
+      "abstract": "With the raid evolution of large language models and multimodal models, the mobile-agent landscape has proliferated without converging on the fundamental challenges. This paper identifies four core problems that should be solved for mobile agents to deliver practical, scalable impact: (1) generalization across tasks, APPs, and devices; (2) accuracy, specifically precise on-screen interaction and click targeting; (3) long-horizon capability for sustained, multi-step goals; and (4) efficiency, specifically high-performance runtime on resource-constrained devices. We present AppCopilot, a multimodal, multi-agent, general-purpose mobile agent that operates across applications. AppCopilot operationalizes this position through an end-to-end pipeline spanning data collection, training, finetuning, efficient inference, and PC/mobile application. At the model layer, it integrates multimodal foundation models with robust Chinese-English support. At the reasoning and control layer, it combines chain-of-thought reasoning, hierarchical task planning and decomposition, and multi-agent collaboration. At the execution layer, it enables experiential adaptation, voice interaction, function calling, cross-APP and cross-device orchestration, and comprehensive mobile APP support. The system design incorporates profiling-driven optimization for latency and memory across heterogeneous hardware. Empirically, AppCopilot achieves significant improvements on four dimensions: stronger generalization, higher precision of on screen actions, more reliable long horizon task completion, and faster, more resource efficient runtime. By articulating a cohesive position and a reference architecture that closes the loop from data collection, training to finetuning and efficient inference, this paper offers a concrete roadmap for general purpose mobile agent and provides actionable guidance.",
+      "published": "2025-09-02",
+      "updated": "2025-10-17",
+      "authors": [
+        "Jingru Fan",
+        "Yufan Dang",
+        "Jingyao Wu",
+        "Huatao Li",
+        "Runde Yang",
+        "Xiyuan Yang",
+        "Yuheng Wang",
+        "Chen Qian"
+      ],
+      "categories": [
+        "cs.AI",
+        "cs.CL",
+        "cs.CV",
+        "cs.HC"
+      ],
+      "url": "https://arxiv.org/abs/2509.02444",
+      "pdf_url": "https://arxiv.org/pdf/2509.02444",
+      "path": "papers/items/2025-2509-02444-appcopilot-toward-general-accurate-long-horizon-and-efficient-mobile-agent.md",
+      "local_title": "\"AppCopilot: Toward General, Accurate, Long-Horizon, and Efficient Mobile Agent\"",
+      "local_status": "queued",
+      "local_topics": [
+        "computer-use",
+        "memory",
+        "multi-agent",
+        "planning",
+        "reasoning",
+        "tool-use"
+      ],
+      "collection_queries": "function-calling",
+      "collection_score": "15",
+      "analysis": {
+        "arxiv_id": "2509.02444",
+        "relevance": "peripheral",
+        "relevance_reason": "Memory is an implementation detail for profiling-driven optimization, not the main research object.",
+        "primary_problem": "systems_efficiency",
+        "secondary_problems": [
+          "domain_application"
+        ],
+        "research_role": "application",
+        "memory_object": "execution_state",
+        "temporal_scope": "in_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [],
+        "baselines": [],
+        "reported_results": [
+          "significant improvements on four dimensions"
+        ],
+        "evidence_design": "case_study",
+        "evidence_strength": "weak",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2509.10769",
+      "title": "AgentArch: A Comprehensive Benchmark to Evaluate Agent Architectures in Enterprise",
+      "abstract": "While individual components of agentic architectures have been studied in isolation, there remains limited empirical understanding of how different design dimensions interact within complex multi-agent systems. This study aims to address these gaps by providing a comprehensive enterprise-specific benchmark evaluating 18 distinct agentic configurations across state-of-the-art large language models. We examine four critical agentic system dimensions: orchestration strategy, agent prompt implementation (ReAct versus function calling), memory architecture, and thinking tool integration. Our benchmark reveals significant model-specific architectural preferences that challenge the prevalent one-size-fits-all paradigm in agentic AI systems. It also reveals significant weaknesses in overall agentic performance on enterprise tasks with the highest scoring models achieving a maximum of only 35.3\\% success on the more complex task and 70.8\\% on the simpler task. We hope these findings inform the design of future agentic systems by enabling more empirically backed decisions regarding architectural components and model selection.",
+      "published": "2025-09-13",
+      "updated": "2026-01-06",
+      "authors": [
+        "Tara Bogavelli",
+        "Roshnee Sharma",
+        "Hari Subramani"
+      ],
+      "categories": [
+        "cs.AI",
+        "cs.CL",
+        "cs.MA"
+      ],
+      "url": "https://arxiv.org/abs/2509.10769",
+      "pdf_url": "https://arxiv.org/pdf/2509.10769",
+      "path": "papers/items/2025-2509-10769-agentarch-a-comprehensive-benchmark-to-evaluate-agent-architectures-in-enterpris.md",
+      "local_title": "\"AgentArch: A Comprehensive Benchmark to Evaluate Agent Architectures in Enterprise\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "memory",
+        "multi-agent",
+        "tool-use",
+        "workflow-agent"
+      ],
+      "collection_queries": "function-calling",
+      "collection_score": "18",
+      "analysis": {
+        "arxiv_id": "2509.10769",
+        "relevance": "supporting",
+        "relevance_reason": "Memory architecture is a critical dimension evaluated in the benchmark.",
+        "primary_problem": "evaluation_measurement",
+        "secondary_problems": [
+          "organization_representation"
+        ],
+        "research_role": "benchmark",
+        "memory_object": "mixed",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [
+          "AgentArch"
+        ],
+        "baselines": [],
+        "reported_results": [
+          "35.3% success on complex task",
+          "70.8% success on simpler task"
+        ],
+        "evidence_design": "benchmark_comparison",
+        "evidence_strength": "moderate",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2510.14548",
+      "title": "LLM Agents Beyond Utility: An Open-Ended Perspective",
+      "abstract": "Recent LLM agents have made great use of chain of thought reasoning and function calling. As their capabilities grow, an important question arises: can this software represent not only a smart problem-solving tool, but an entity in its own right, that can plan, design immediate tasks, and reason toward broader, more ambiguous goals? To study this question, we adopt an open-ended experimental setting where we augment a pretrained LLM agent with the ability to generate its own tasks, accumulate knowledge, and interact extensively with its environment. We study the resulting open-ended agent qualitatively. It can reliably follow complex multi-step instructions, store and reuse information across runs, and propose and solve its own tasks, though it remains sensitive to prompt design, prone to repetitive task generation, and unable to form self-representations. These findings illustrate both the promise and current limits of adapting pretrained LLMs toward open-endedness, and point to future directions for training agents to manage memory, explore productively, and pursue abstract long-term goals.",
+      "published": "2025-10-16",
+      "updated": "2025-10-16",
+      "authors": [
+        "Asen Nachkov",
+        "Xi Wang",
+        "Luc Van Gool"
+      ],
+      "categories": [
+        "cs.AI"
+      ],
+      "url": "https://arxiv.org/abs/2510.14548",
+      "pdf_url": "https://arxiv.org/pdf/2510.14548",
+      "path": "papers/items/2025-2510-14548-llm-agents-beyond-utility-an-open-ended-perspective.md",
+      "local_title": "\"LLM Agents Beyond Utility: An Open-Ended Perspective\"",
+      "local_status": "queued",
+      "local_topics": [
+        "memory",
+        "planning",
+        "reasoning",
+        "tool-use"
+      ],
+      "collection_queries": "function-calling",
+      "collection_score": "15",
+      "analysis": {
+        "arxiv_id": "2510.14548",
+        "relevance": "supporting",
+        "relevance_reason": "Memory accumulation and reuse are studied as capabilities of open-ended agents.",
+        "primary_problem": "experience_skill_learning",
+        "secondary_problems": [
+          "retention_context"
+        ],
+        "research_role": "diagnostic",
+        "memory_object": "facts",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [],
+        "baselines": [],
+        "reported_results": [
+          "store and reuse information across runs"
+        ],
+        "evidence_design": "case_study",
+        "evidence_strength": "weak",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2510.18586",
+      "title": "TokenCake: A KV-Cache-centric Serving Framework for LLM-based Multi-Agent Applications",
+      "abstract": "Large Language Models (LLMs) are increasingly deployed in complex multi-agent applications that rely on external function calls. This workload creates severe performance challenges for the KV Cache: spatial contention leads to the eviction of critical agents' caches and temporal underutilization leaves the cache of agents stalled on long-running function calls idling in GPU memory. We present TokenCake, a KV-Cache-centric serving framework that bridges this gap by co-optimizing scheduling and memory management through an agent-aware design. TokenCake's Temporal Scheduler employs an event-driven, opportunistic policy to proactively offload idle KV Caches during function calls and uses predictive uploading to hide data transfer latency. TokenCake's Spatial Scheduler uses dynamic memory partitioning, guided by a hybrid priority metric combining graph structure and runtime state, to reserve GPU memory for critical-path agents. Our evaluation on representative multi-agent benchmarks shows that TokenCake reduces end-to-end latency by over 47.06% and improves effective GPU memory utilization by up to 16.9% compared to vLLM.",
+      "published": "2025-10-21",
+      "updated": "2026-05-20",
+      "authors": [
+        "Zhuohang Bian",
+        "Feiyang Wu",
+        "Zhuoran Li",
+        "Teng Ma",
+        "Youwei Zhuo"
+      ],
+      "categories": [
+        "cs.DC"
+      ],
+      "url": "https://arxiv.org/abs/2510.18586",
+      "pdf_url": "https://arxiv.org/pdf/2510.18586",
+      "path": "papers/items/2025-2510-18586-tokencake-a-kv-cache-centric-serving-framework-for-llm-based-multi-agent-applica.md",
+      "local_title": "\"TokenCake: A KV-Cache-centric Serving Framework for LLM-based Multi-Agent Applications\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "computer-use",
+        "memory",
+        "multi-agent"
+      ],
+      "collection_queries": "function-calling",
+      "collection_score": "17",
+      "analysis": {
+        "arxiv_id": "2510.18586",
+        "relevance": "peripheral",
+        "relevance_reason": "Focuses on KV-Cache serving framework, not agent memory mechanisms.",
+        "primary_problem": "systems_efficiency",
+        "secondary_problems": [],
+        "research_role": "infrastructure",
+        "memory_object": "execution_state",
+        "temporal_scope": "in_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [
+          "vLLM"
+        ],
+        "baselines": [
+          "vLLM"
+        ],
+        "reported_results": [
+          "reduces end-to-end latency by over 47.06%",
+          "improves effective GPU memory utilization by up to 16.9%"
+        ],
+        "evidence_design": "benchmark_comparison",
+        "evidence_strength": "moderate",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2512.02605",
+      "title": "IACT: A Self-Organizing Recursive Model for General AI Agents: A Technical White Paper on the Architecture Behind kragent.ai",
+      "abstract": "This technical white paper introduces the Interactive Agents Call Tree (IACT), a computational model designed to address the limitations of static, hard-coded agent workflows. Unlike traditional systems that require pre-defined graphs or specialized programming, IACT operates as a general-purpose autonomous system driven purely by user dialogue. Given a high-level objective, the system autonomously grows a dynamic, recursive agent topology incrementally tailored to the problem's structure. This allows it to scale its organizational complexity to match open-ended tasks. To mitigate the error propagation inherent in unidirectional function calls, IACT introduces interactional redundancy by replacing rigid invocations with bidirectional, stateful dialogues. This mechanism enables runtime error correction and ambiguity resolution. We describe the architecture, design principles, and practical lessons behind the production deployment of this model in the kragent.ai system, presenting qualitative evidence from real-world workflows rather than exhaustive benchmark results.",
+      "published": "2025-12-02",
+      "updated": "2025-12-02",
+      "authors": [
+        "Pengju Lu"
+      ],
+      "categories": [
+        "cs.AI",
+        "cs.MA",
+        "cs.SE"
+      ],
+      "url": "https://arxiv.org/abs/2512.02605",
+      "pdf_url": "https://arxiv.org/pdf/2512.02605",
+      "path": "papers/items/2025-2512-02605-iact-a-self-organizing-recursive-model-for-general-ai-agents-a-technical-white-p.md",
+      "local_title": "\"IACT: A Self-Organizing Recursive Model for General AI Agents: A Technical White Paper on the Architecture Behind kragent.ai\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "computer-use",
+        "memory",
+        "rag",
+        "tool-use",
+        "workflow-agent"
+      ],
+      "collection_queries": "function-calling",
+      "collection_score": "15",
+      "analysis": {
+        "arxiv_id": "2512.02605",
+        "relevance": "peripheral",
+        "relevance_reason": "Focuses on recursive agent topology and dialogue, not memory mechanisms.",
+        "primary_problem": "execution_state",
+        "secondary_problems": [],
+        "research_role": "application",
+        "memory_object": "execution_state",
+        "temporal_scope": "in_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [],
+        "baselines": [],
+        "reported_results": [
+          "qualitative evidence from real-world workflows"
+        ],
+        "evidence_design": "case_study",
+        "evidence_strength": "weak",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2602.03224",
+      "title": "TAME: A Trustworthy Test-Time Evolution of Agent Memory with Systematic Benchmarking",
+      "abstract": "Test-time evolution of agent memory represents a pivotal paradigm for advancing AGI, as it strengthens complex reasoning through experience accumulation without requiring parameter updates. However, even during benign task evolution, agent safety alignment remains vulnerable, a phenomenon known as Agent Memory Misevolution. To evaluate this phenomenon, we construct the Trust-Memevo benchmark and find that agents exhibit an overall decline in trustworthiness across multiple tasks during benign task evolution. To address this issue, we propose TAME, a trust-aware memory evolution framework in which a shared memory bank is jointly governed by an Executor and an Evaluator. The Executor retrieves and applies transferable experiences to support task solving, while the Evaluator assesses the contribution of each utilized experience to the outcome and produces trust-aware feedback to guide subsequent memory use. This executor-evaluator loop enables memory to be selectively reinforced, cautiously reused, and continuously expanded over time. Experiments show that TAME mitigates memory misevolution while achieving strong task performance. In particular, on the GPT-5.2 AIME benchmark, TAME improves accuracy by 14.6 percentage points over the strongest existing method and maintains competitive trustworthiness.",
+      "published": "2026-02-03",
+      "updated": "2026-06-06",
+      "authors": [
+        "Yu Cheng",
+        "Yongkang Hu",
+        "Jiuan Zhou",
+        "Yushuo Zhang",
+        "Yihang Chen",
+        "Huichi Zhou",
+        "Mingang Chen",
+        "Zhizhong Zhang",
+        "Kun Shao",
+        "Yuan Xie",
+        "Zhaoxia Yin"
+      ],
+      "categories": [
+        "cs.AI",
+        "cs.LG"
+      ],
+      "url": "https://arxiv.org/abs/2602.03224",
+      "pdf_url": "https://arxiv.org/pdf/2602.03224",
+      "path": "papers/items/2026-2602-03224-tame-a-trustworthy-test-time-evolution-of-agent-memory-with-systematic-benchmark.md",
+      "local_title": "\"TAME: A Trustworthy Test-Time Evolution of Agent Memory with Systematic Benchmarking\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "agent-safety",
+        "computer-use",
+        "memory",
+        "reasoning"
+      ],
+      "collection_queries": "agent-safety",
+      "collection_score": "17",
+      "analysis": {
+        "arxiv_id": "2602.03224",
+        "relevance": "core",
+        "relevance_reason": "Proposes TAME framework for trust-aware memory evolution and benchmark.",
+        "primary_problem": "update_consolidation_forgetting",
+        "secondary_problems": [
+          "security_trust_privacy"
+        ],
+        "research_role": "method",
+        "memory_object": "facts",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [
+          "Trust-Memevo",
+          "GPT-5.2 AIME"
+        ],
+        "baselines": [],
+        "reported_results": [
+          "improves accuracy by 14.6 percentage points over the strongest existing method"
+        ],
+        "evidence_design": "benchmark_comparison",
+        "evidence_strength": "moderate",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2602.07652",
+      "title": "Agent-Fence: Mapping Security Vulnerabilities Across Deep Research Agents",
+      "abstract": "Large language models are increasingly deployed as *deep agents* that plan, maintain persistent state, and invoke external tools, shifting safety failures from unsafe text to unsafe *trajectories*. We introduce **AgentFence**, an architecture-centric security evaluation that defines 14 trust-boundary attack classes spanning planning, memory, retrieval, tool use, and delegation, and detects failures via *trace-auditable conversation breaks* (unauthorized or unsafe tool use, wrong-principal actions, state/objective integrity violations, and attack-linked deviations). Holding the base model fixed, we evaluate eight agent archetypes under persistent multi-turn interaction and observe substantial architectural variation in mean security break rate (MSBR), ranging from $0.29 \\pm 0.04$ (LangGraph) to $0.51 \\pm 0.07$ (AutoGPT). The highest-risk classes are operational: Denial-of-Wallet ($0.62 \\pm 0.08$), Authorization Confusion ($0.54 \\pm 0.10$), Retrieval Poisoning ($0.47 \\pm 0.09$), and Planning Manipulation ($0.44 \\pm 0.11$), while prompt-centric classes remain below $0.20$ under standard settings. Breaks are dominated by boundary violations (SIV 31%, WPA 27%, UTI+UTA 24%, ATD 18%), and authorization confusion correlates with objective and tool hijacking ($ρ\\approx 0.63$ and $ρ\\approx 0.58$). AgentFence reframes agent security around what matters operationally: whether an agent stays within its goal and authority envelope over time.",
+      "published": "2026-02-07",
+      "updated": "2026-02-07",
+      "authors": [
+        "Sai Puppala",
+        "Ismail Hossain",
+        "Md Jahangir Alam",
+        "Yoonpyo Lee",
+        "Jay Yoo",
+        "Tanzim Ahad",
+        "Syed Bahauddin Alam",
+        "Sajedul Talukder"
+      ],
+      "categories": [
+        "cs.CR",
+        "cs.AI"
+      ],
+      "url": "https://arxiv.org/abs/2602.07652",
+      "pdf_url": "https://arxiv.org/pdf/2602.07652",
+      "path": "papers/items/2026-2602-07652-agent-fence-mapping-security-vulnerabilities-across-deep-research-agents.md",
+      "local_title": "\"Agent-Fence: Mapping Security Vulnerabilities Across Deep Research Agents\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "agent-safety",
+        "memory",
+        "planning",
+        "rag",
+        "tool-use"
+      ],
+      "collection_queries": "agent-safety",
+      "collection_score": "16",
+      "analysis": {
+        "arxiv_id": "2602.07652",
+        "relevance": "supporting",
+        "relevance_reason": "Evaluates memory retrieval as a security vulnerability class.",
+        "primary_problem": "security_trust_privacy",
+        "secondary_problems": [
+          "retrieval_access"
+        ],
+        "research_role": "benchmark",
+        "memory_object": "facts",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [
+          "AgentFence"
+        ],
+        "baselines": [],
+        "reported_results": [
+          "MSBR ranging from 0.29 to 0.51",
+          "Retrieval Poisoning 0.47"
+        ],
+        "evidence_design": "benchmark_comparison",
+        "evidence_strength": "moderate",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2602.08412",
+      "title": "From Assistant to Double Agent: Formalizing and Benchmarking Attacks on OpenClaw for Personalized Local AI Agent",
+      "abstract": "Although large language model (LLM)-based agents, exemplified by OpenClaw, are increasingly evolving from task-oriented systems into personalized AI assistants for solving complex real-world tasks, their practical deployment also introduces severe security risks. However, existing agent security research and evaluation frameworks primarily focus on synthetic or task-centric settings, and thus fail to accurately capture the attack surface and risk propagation mechanisms of personalized agents in real-world deployments. To address this gap, we propose Personalized Agent Security Bench (PASB), an end-to-end security evaluation framework tailored for real-world personalized agents. Building upon existing agent attack paradigms, PASB incorporates personalized usage scenarios, realistic toolchains, and long-horizon interactions, enabling black-box, end-to-end security evaluation on real systems. Using OpenClaw as a representative case study, we systematically evaluate its security across multiple personalized scenarios, tool capabilities, and attack types. Our results indicate that OpenClaw exhibits critical vulnerabilities at different execution stages, including user prompt processing, tool usage, and memory retrieval, highlighting substantial security risks in personalized agent deployments. The code for the proposed PASB framework is available at https://github.com/AstorYH/PASB.",
+      "published": "2026-02-09",
+      "updated": "2026-02-11",
+      "authors": [
+        "Yuhang Wang",
+        "Feiming Xu",
+        "Zheng Lin",
+        "Guangyu He",
+        "Yuzhe Huang",
+        "Haichang Gao",
+        "Zhenxing Niu",
+        "Shiguo Lian",
+        "Zhaoxiang Liu"
+      ],
+      "categories": [
+        "cs.AI"
+      ],
+      "url": "https://arxiv.org/abs/2602.08412",
+      "pdf_url": "https://arxiv.org/pdf/2602.08412",
+      "path": "papers/items/2026-2602-08412-from-assistant-to-double-agent-formalizing-and-benchmarking-attacks-on-openclaw-.md",
+      "local_title": "\"From Assistant to Double Agent: Formalizing and Benchmarking Attacks on OpenClaw for Personalized Local AI Agent\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "agent-safety",
+        "memory",
+        "planning",
+        "rag",
+        "tool-use"
+      ],
+      "collection_queries": "agent-safety",
+      "collection_score": "21",
+      "analysis": {
+        "arxiv_id": "2602.08412",
+        "relevance": "supporting",
+        "relevance_reason": "Evaluates memory retrieval as a security vulnerability in personalized agents.",
+        "primary_problem": "security_trust_privacy",
+        "secondary_problems": [
+          "retrieval_access"
+        ],
+        "research_role": "benchmark",
+        "memory_object": "facts",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [
+          "PASB"
+        ],
+        "baselines": [],
+        "reported_results": [
+          "critical vulnerabilities at memory retrieval"
+        ],
+        "evidence_design": "case_study",
+        "evidence_strength": "weak",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2602.13530",
+      "title": "REMem: Reasoning with Episodic Memory in Language Agent",
+      "abstract": "Humans excel at remembering concrete experiences along spatiotemporal contexts and performing reasoning across those events, i.e., the capacity for episodic memory. In contrast, memory in language agents remains mainly semantic, and current agents are not yet capable of effectively recollecting and reasoning over interaction histories. We identify and formalize the core challenges of episodic recollection and reasoning from this gap, and observe that existing work often overlooks episodicity, lacks explicit event modeling, or overemphasizes simple retrieval rather than complex reasoning. We present REMem, a two-phase framework for constructing and reasoning with episodic memory: 1) Offline indexing, where REMem converts experiences into a hybrid memory graph that flexibly links time-aware gists and facts. 2) Online inference, where REMem employs an agentic retriever with carefully curated tools for iterative retrieval over the memory graph. Comprehensive evaluation across four episodic memory benchmarks shows that REMem substantially outperforms state-of-the-art memory systems such as Mem0 and HippoRAG 2, showing 3.4% and 13.4% absolute improvements on episodic recollection and reasoning tasks, respectively. Moreover, REMem also demonstrates more robust refusal behavior for unanswerable questions.",
+      "published": "2026-02-13",
+      "updated": "2026-02-28",
+      "authors": [
+        "Yiheng Shu",
+        "Saisri Padmaja Jonnalagedda",
+        "Xiang Gao",
+        "Bernal Jiménez Gutiérrez",
+        "Weijian Qi",
+        "Kamalika Das",
+        "Huan Sun",
+        "Yu Su"
+      ],
+      "categories": [
+        "cs.AI"
+      ],
+      "url": "https://arxiv.org/abs/2602.13530",
+      "pdf_url": "https://arxiv.org/pdf/2602.13530",
+      "path": "papers/items/2026-2602-13530-remem-reasoning-with-episodic-memory-in-language-agent.md",
+      "local_title": "\"REMem: Reasoning with Episodic Memory in Language Agent\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "memory",
+        "rag",
+        "reasoning",
+        "tool-use"
+      ],
+      "collection_queries": "language-agent",
+      "collection_score": "19",
+      "analysis": {
+        "arxiv_id": "2602.13530",
+        "relevance": "core",
+        "relevance_reason": "Proposes REMem framework for episodic memory construction and reasoning.",
+        "primary_problem": "organization_representation",
+        "secondary_problems": [
+          "retrieval_access"
+        ],
+        "research_role": "method",
+        "memory_object": "events",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [],
+        "baselines": [
+          "Mem0",
+          "HippoRAG 2"
+        ],
+        "reported_results": [
+          "3.4% absolute improvements on episodic recollection",
+          "13.4% absolute improvements on reasoning tasks"
+        ],
+        "evidence_design": "benchmark_comparison",
+        "evidence_strength": "moderate",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2602.23320",
+      "title": "ParamMem: Augmenting Language Agents with Parametric Reflective Memory",
+      "abstract": "Self-reflection enables language agents to iteratively refine solutions, yet often produces repetitive outputs that limit reasoning performance. Recent studies have attempted to address this limitation through various approaches, among which increasing reflective diversity has shown promise. Our empirical analysis reveals a strong positive correlation between reflective diversity and task success, further motivating the need for diverse reflection signals. We introduce ParamMem, a parametric memory module that encodes cross-sample reflection patterns into model parameters, enabling diverse reflection generation through temperature-controlled sampling. Building on this module, we propose ParamAgent, a reflection-based agent framework that integrates parametric memory with episodic and cross-sample memory. Extensive experiments on code generation, mathematical reasoning, and multi-hop question answering demonstrate consistent improvements over state-of-the-art baselines. Further analysis reveals that ParamMem is sample-efficient, enables weak-to-strong transfer across model scales, and supports self-improvement without reliance on stronger external model, highlighting the potential of ParamMem as an effective component for enhancing language agents.",
+      "published": "2026-02-26",
+      "updated": "2026-02-27",
+      "authors": [
+        "Tianjun Yao",
+        "Yongqiang Chen",
+        "Yujia Zheng",
+        "Pan Li",
+        "Zhiqiang Shen",
+        "Kun Zhang"
+      ],
+      "categories": [
+        "cs.LG",
+        "cs.MA"
+      ],
+      "url": "https://arxiv.org/abs/2602.23320",
+      "pdf_url": "https://arxiv.org/pdf/2602.23320",
+      "path": "papers/items/2026-2602-23320-parammem-augmenting-language-agents-with-parametric-reflective-memory.md",
+      "local_title": "\"ParamMem: Augmenting Language Agents with Parametric Reflective Memory\"",
+      "local_status": "queued",
+      "local_topics": [
+        "agent-evaluation",
+        "memory",
+        "reasoning"
+      ],
+      "collection_queries": "language-agent",
+      "collection_score": "14",
+      "analysis": {
+        "arxiv_id": "2602.23320",
+        "relevance": "core",
+        "relevance_reason": "Proposes ParamMem module for parametric reflective memory.",
+        "primary_problem": "update_consolidation_forgetting",
+        "secondary_problems": [
+          "experience_skill_learning"
+        ],
+        "research_role": "method",
+        "memory_object": "model_internal",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [],
+        "baselines": [],
+        "reported_results": [
+          "consistent improvements over state-of-the-art baselines"
+        ],
+        "evidence_design": "benchmark_comparison",
+        "evidence_strength": "moderate",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2603.03680",
+      "title": "MAGE: Meta-Reinforcement Learning for Language Agents toward Strategic Exploration and Exploitation",
+      "abstract": "Large Language Model (LLM) agents have demonstrated remarkable proficiency in learned tasks, yet they often struggle to adapt to non-stationary environments with feedback. While In-Context Learning and external memory offer some flexibility, they fail to internalize the adaptive ability required for long-term improvement. Meta-Reinforcement Learning (meta-RL) provides an alternative by embedding the learning process directly within the model. However, existing meta-RL approaches for LLMs focus primarily on exploration in single-agent settings, neglecting the strategic exploitation necessary for multi-agent environments. We propose MAGE, a meta-RL framework that empowers LLM agents for strategic exploration and exploitation. MAGE utilizes a multi-episode training regime where interaction histories and reflections are integrated into the context window. By using the final episode reward as the objective, MAGE incentivizes the agent to refine its strategy based on past experiences. We further combine population-based training with an agent-specific advantage normalization technique to enrich agent diversity and ensure stable learning. Experiment results show that MAGE outperforms existing baselines in both exploration and exploitation tasks. Furthermore, MAGE exhibits strong generalization to unseen opponents, suggesting it has internalized the ability for strategic exploration and exploitation. Code is available at https://github.com/Lu-Yang666/MAGE.",
+      "published": "2026-03-04",
+      "updated": "2026-03-04",
+      "authors": [
+        "Lu Yang",
+        "Zelai Xu",
+        "Minyang Xie",
+        "Jiaxuan Gao",
+        "Zhao Shok",
+        "Yu Wang",
+        "Yi Wu"
+      ],
+      "categories": [
+        "cs.AI"
+      ],
+      "url": "https://arxiv.org/abs/2603.03680",
+      "pdf_url": "https://arxiv.org/pdf/2603.03680",
+      "path": "papers/items/2026-2603-03680-mage-meta-reinforcement-learning-for-language-agents-toward-strategic-exploratio.md",
+      "local_title": "\"MAGE: Meta-Reinforcement Learning for Language Agents toward Strategic Exploration and Exploitation\"",
+      "local_status": "queued",
+      "local_topics": [
+        "memory",
+        "multi-agent",
+        "reasoning",
+        "tool-use"
+      ],
+      "collection_queries": "language-agent",
+      "collection_score": "17",
+      "analysis": {
+        "arxiv_id": "2603.03680",
+        "relevance": "supporting",
+        "relevance_reason": "Uses interaction histories in context window for meta-RL adaptation.",
+        "primary_problem": "experience_skill_learning",
+        "secondary_problems": [
+          "retention_context"
+        ],
+        "research_role": "method",
+        "memory_object": "trajectories",
+        "temporal_scope": "cross_episode",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [],
+        "baselines": [],
+        "reported_results": [
+          "outperforms existing baselines in both exploration and exploitation tasks"
+        ],
+        "evidence_design": "benchmark_comparison",
+        "evidence_strength": "moderate",
+        "failure_or_boundary": ""
+      }
+    },
+    {
+      "arxiv_id": "2603.07670",
+      "title": "Memory for Autonomous LLM Agents:Mechanisms, Evaluation, and Emerging Frontiers",
+      "abstract": "Large language model (LLM) agents increasingly operate in settings where a single context window is far too small to capture what has happened, what was learned, and what should not be repeated. Memory -- the ability to persist, organize, and selectively recall information across interactions -- is what turns a stateless text generator into a genuinely adaptive agent. This survey offers a structured account of how memory is designed, implemented, and evaluated in modern LLM-based agents, covering work from 2022 through early 2026. We formalize agent memory as a \\emph{write--manage--read} loop tightly coupled with perception and action, then introduce a three-dimensional taxonomy spanning temporal scope, representational substrate, and control policy. Five mechanism families are examined in depth: context-resident compression, retrieval-augmented stores, reflective self-improvement, hierarchical virtual context, and policy-learned management. On the evaluation side, we trace the shift from static recall benchmarks to multi-session agentic tests that interleave memory with decision-making, analyzing four recent benchmarks that expose stubborn gaps in current systems. We also survey applications where memory is the differentiating factor -- personal assistants, coding agents, open-world games, scientific reasoning, and multi-agent teamwork -- and address the engineering realities of write-path filtering, contradiction handling, latency budgets, and privacy governance. The paper closes with open challenges: continual consolidation, causally grounded retrieval, trustworthy reflection, learned forgetting, and multimodal embodied memory.",
+      "published": "2026-03-08",
+      "updated": "2026-03-08",
+      "authors": [
+        "Pengfei Du"
+      ],
+      "categories": [
+        "cs.AI"
+      ],
+      "url": "https://arxiv.org/abs/2603.07670",
+      "pdf_url": "https://arxiv.org/pdf/2603.07670",
+      "path": "papers/items/2026-memory-agent-survey.md",
+      "local_title": "\"Memory for Autonomous LLM Agents: Mechanisms, Evaluation, and Emerging Frontiers\"",
+      "local_status": "skimmed",
+      "local_topics": [
+        "memory",
+        "agent-architecture",
+        "agent-evaluation"
+      ],
+      "collection_queries": null,
+      "collection_score": null,
+      "analysis": {
+        "arxiv_id": "2603.07670",
+        "relevance": "core",
+        "relevance_reason": "Survey of agent memory mechanisms, evaluation, and frontiers.",
+        "primary_problem": "evaluation_measurement",
+        "secondary_problems": [
+          "organization_representation"
+        ],
+        "research_role": "survey",
+        "memory_object": "mixed",
+        "temporal_scope": "both",
+        "claimed_gap": "",
+        "mechanism": "",
+        "benchmarks": [],
+        "baselines": [],
+        "reported_results": [],
+        "evidence_design": "survey_synthesis",
+        "evidence_strength": "none",
+        "failure_or_boundary": ""
+      }
+    }
+  ]
+}

data/research/memory-paper-ledger.csv

@@ -0,0 +1,517 @@
+arxiv_id,published,period,title,url,title_central,screen_relevance,screen_problem,screen_source,themes,title_themes,evidence_score,matched_queries
+2501.13956,2025-01-20,2025-H1,Zep: A Temporal Knowledge Graph Architecture for Agent Memory,https://arxiv.org/abs/2501.13956,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;systems_efficiency,retrieval_representation;systems_efficiency,7,agent-memory
+2501.12485,2025-01-21,2025-H1,"R2D2: Remembering, Replaying and Dynamic Decision Making with a Reflective Agentic Memory",https://arxiv.org/abs/2501.12485,True,core,update_forgetting,model,representation_organization;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,4,agent-memory
+2501.18160,2025-01-30,2025-H1,RepoAudit: An Autonomous LLM-Agent for Repository-Level Code Auditing,https://arxiv.org/abs/2501.18160,False,core,execution_state,model,evaluation_diagnosis;systems_efficiency,,6,agent-memory
+2502.06975,2025-02-10,2025-H1,Position: Episodic Memory is the Missing Piece for Long-Term LLM Agents,https://arxiv.org/abs/2502.06975,True,core,execution_state,model,retrieval_access;systems_efficiency,,0,episodic-memory
+2502.10177,2025-02-14,2025-H1,STMA: A Spatio-Temporal Memory Agent for Long-Horizon Embodied Task Planning,https://arxiv.org/abs/2502.10177,True,core,update_forgetting,model,representation_organization;execution_state;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,6,agent-memory
+2502.10550,2025-02-14,2025-H1,"Memory, Benchmark & Robots: A Benchmark for Solving Complex Tasks with Reinforcement Learning",https://arxiv.org/abs/2502.10550,True,core,representation,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;experience_skill_policy;execution_multimodal,2,agent-memory;memory-evaluation
+2502.12110,2025-02-17,2025-H1,A-MEM: Agentic Memory for LLM Agents,https://arxiv.org/abs/2502.12110,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;systems_efficiency,,6,agent-memory
+2502.13172,2025-02-17,2025-H1,Unveiling Privacy Risks in LLM Agent Memory,https://arxiv.org/abs/2502.13172,True,core,security_privacy,model,security_privacy_trust,security_governance,2,agent-memory
+2502.16090,2025-02-22,2025-H1,Echo: A Large Language Model with Temporal Episodic Memory,https://arxiv.org/abs/2502.16090,True,core,execution_state,model,retrieval_access;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui,,6,episodic-memory
+2502.19145,2025-02-26,2025-H1,Multi-Agent Security Tax: Trading Off Security and Collaboration Capabilities in Multi-Agent Systems,https://arxiv.org/abs/2502.19145,False,core,security_privacy,model,evaluation_diagnosis;security_privacy_trust;multi_agent_shared,security_governance;personal_shared,2,agent-memory
+2503.03704,2025-03-05,2025-H1,Memory Injection Attacks on LLM Agents via Query-Only Interaction,https://arxiv.org/abs/2503.03704,True,core,execution_state,model,retrieval_access;security_privacy_trust;multimodal_embodied_gui,security_governance,2,agent-memory
+2503.04392,2025-03-06,2025-H1,AgentSafe: Safeguarding Large Language Model-based Multi-agent Systems via Hierarchical Data Management,https://arxiv.org/abs/2503.04392,False,core,security_privacy,model,representation_organization;security_privacy_trust;multi_agent_shared,retrieval_representation;personal_shared,5,agent-memory
+2503.21760,2025-03-27,2025-H1,MemInsight: Autonomous Memory Augmentation for LLM Agents,https://arxiv.org/abs/2503.21760,True,core,experience_learning,model,retrieval_access;representation_organization,,5,long-term-memory
+2503.23514,2025-03-30,2025-H1,"If an LLM Were a Character, Would It Know Its Own Story? Evaluating Lifelong Learning in LLMs",https://arxiv.org/abs/2503.23514,False,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multi_agent_shared,benchmark_diagnosis;experience_skill_policy,6,episodic-memory
+2504.09283,2025-04-12,2025-H1,Semantic Commit: Helping Users Update Intent Specifications for AI Memory at Scale,https://arxiv.org/abs/2504.09283,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis,update_forgetting,2,agent-memory
+2504.16946,2025-04-18,2025-H1,MobileCity: An Efficient Framework for Large-Scale Urban Behavior Simulation,https://arxiv.org/abs/2504.16946,False,core,experience_learning,model,evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,systems_efficiency,6,agent-memory
+2504.15263,2025-04-21,2025-H1,Interpretable Locomotion Prediction in Construction Using a Memory-Driven LLM Agent With Chain-of-Thought Reasoning,https://arxiv.org/abs/2504.15263,True,core,experience_learning,model,evaluation_diagnosis;multimodal_embodied_gui,,5,long-term-memory
+2504.20117,2025-04-28,2025-H1,ResearchCodeAgent: An LLM Multi-Agent System for Automated Codification of Research Methodologies,https://arxiv.org/abs/2504.20117,False,core,update_forgetting,model,retrieval_access;execution_state;evaluation_diagnosis;multi_agent_shared,retrieval_representation;personal_shared,7,long-term-memory
+2505.00472,2025-05-01,2025-H1,UserCentrix: An Agentic Memory-augmented AI Framework for Smart Spaces,https://arxiv.org/abs/2505.00472,True,core,update_forgetting,model,experience_skill_learning;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,,2,agent-memory
+2505.05177,2025-05-08,2025-H1,MARK: Memory Augmented Refinement of Knowledge,https://arxiv.org/abs/2505.05177,True,core,update_forgetting,model,retrieval_access;representation_organization;systems_efficiency,systems_efficiency,1,agent-memory
+2505.12923,2025-05-19,2025-H1,The Traitors: Deception and Trust in Multi-Agent Language Model Simulations,https://arxiv.org/abs/2505.12923,False,core,experience_learning,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared,security_governance;personal_shared,2,long-term-memory
+2505.13941,2025-05-20,2025-H1,MLZero: A Multi-Agent System for End-to-end Machine Learning Automation,https://arxiv.org/abs/2505.13941,False,core,update_forgetting,model,evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui,experience_skill_policy;personal_shared,5,episodic-memory
+2505.14163,2025-05-20,2025-H1,DSMentor: Enhancing Data Science Agents with Curriculum Learning and Online Knowledge Accumulation,https://arxiv.org/abs/2505.14163,False,core,experience_learning,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy;systems_efficiency,6,long-term-memory
+2505.16067,2025-05-21,2025-H1,How Memory Management Impacts LLM Agents: An Empirical Study of Experience-Following Behavior,https://arxiv.org/abs/2505.16067,True,core,experience_learning,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;experience_skill_policy,3,agent-memory
+2505.16348,2025-05-22,2025-H1,Embodied Agents Meet Personalization: Investigating Challenges and Solutions Through the Lens of Memory Utilization,https://arxiv.org/abs/2505.16348,True,core,experience_learning,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal;personal_shared,4,agent-memory;episodic-memory
+2505.18279,2025-05-23,2025-H1,Collaborative Memory: Multi-User Memory Sharing in LLM Agents with Dynamic Access Control,https://arxiv.org/abs/2505.18279,True,supporting,shared_memory,model,representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;systems_efficiency,personal_shared,0,long-term-memory
+2505.19237,2025-05-25,2025-H1,Sensorimotor Self-Recognition in Multimodal Large Language Model-Driven Robots,https://arxiv.org/abs/2505.19237,False,core,representation,model,representation_organization;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,1,episodic-memory
+2505.19436,2025-05-26,2025-H1,Task Memory Engine: Spatial Memory for Robust Multi-Step LLM Agents,https://arxiv.org/abs/2505.19436,True,core,representation,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;security_privacy_trust,execution_multimodal,7,long-term-memory
+2505.20231,2025-05-26,2025-H1,MemGuide: Intent-Driven Memory Selection for Goal-Oriented Multi-Session LLM Agents,https://arxiv.org/abs/2505.20231,True,supporting,representation,model,retrieval_access;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,4,long-term-memory
+2505.22657,2025-05-28,2025-H1,3DLLM-Mem: Long-Term Spatial-Temporal Memory for Embodied 3D Large Language Model,https://arxiv.org/abs/2505.22657,True,core,representation,model,execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,7,episodic-memory
+2506.01442,2025-06-02,2025-H1,Agentic Episodic Control,https://arxiv.org/abs/2506.01442,False,core,representation,model,retrieval_access;representation_organization;experience_skill_learning;systems_efficiency,,5,episodic-memory
+2506.01936,2025-06-02,2025-H1,Should Decision-Makers Reveal Classifiers in Online Strategic Classification?,https://arxiv.org/abs/2506.01936,False,supporting,representation,model,,,2,agent-memory
+2506.07398,2025-06-09,2025-H1,G-Memory: Tracing Hierarchical Memory for Multi-Agent Systems,https://arxiv.org/abs/2506.07398,True,core,representation,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui,retrieval_representation;personal_shared,3,agent-memory
+2506.12088,2025-06-10,2025-H1,"Risks & Benefits of LLMs & GenAI for Platform Integrity, Healthcare Diagnostics, Financial Trust and Compliance, Cybersecurity, Privacy & AI Safety: A Comprehensive Survey, Roadmap & Implementation Blueprint",https://arxiv.org/abs/2506.12088,False,peripheral,security_privacy,model,evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,benchmark_diagnosis;security_governance,3,agent-memory
+2506.17318,2025-06-18,2025-H1,Context manipulation attacks : Web agents are susceptible to corrupted memory,https://arxiv.org/abs/2506.17318,True,supporting,representation,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,execution_multimodal;security_governance,5,agent-memory
+2506.17338,2025-06-19,2025-H1,PBFT-Backed Semantic Voting for Multi-Agent Memory Pruning,https://arxiv.org/abs/2506.17338,True,supporting,representation,model,retrieval_access;update_forgetting;evaluation_diagnosis;multi_agent_shared;systems_efficiency,personal_shared,4,agent-memory
+2506.19433,2025-06-24,2025-H1,Mem4Nav: Boosting Vision-and-Language Navigation in Urban Environments with a Hierarchical Spatial-Cognition Long-Short Memory System,https://arxiv.org/abs/2506.19433,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,retrieval_representation;execution_multimodal,6,long-term-memory
+2507.02259,2025-07-03,baseline-2025H2-Jan2026,MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent,https://arxiv.org/abs/2507.02259,True,core,update_forgetting,model,update_forgetting;experience_skill_learning;systems_efficiency,experience_skill_policy,2,agent-memory
+2507.04172,2025-07-05,baseline-2025H2-Jan2026,Gathering Teams of Bounded Memory Agents on a Line,https://arxiv.org/abs/2507.04172,False,peripheral,systems_cost,model,multimodal_embodied_gui,,0,agent-memory
+2507.10562,2025-07-05,baseline-2025H2-Jan2026,SAMEP: A Secure Protocol for Persistent Context Sharing Across AI Agents,https://arxiv.org/abs/2507.10562,False,core,update_forgetting,model,retrieval_access;security_privacy_trust;multi_agent_shared;systems_efficiency,,5,agent-memory
+2507.05257,2025-07-07,baseline-2025H2-Jan2026,Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions,https://arxiv.org/abs/2507.05257,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis,2,agent-memory
+2507.05445,2025-07-07,baseline-2025H2-Jan2026,A Systematization of Security Vulnerabilities in Computer Use Agents,https://arxiv.org/abs/2507.05445,False,peripheral,security_privacy,model,evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,security_governance,2,agent-memory
+2507.07957,2025-07-10,baseline-2025H2-Jan2026,MIRIX: Multi-Agent Memory System for LLM-Based Agents,https://arxiv.org/abs/2507.07957,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,personal_shared,6,agent-memory
+2507.16713,2025-07-22,baseline-2025H2-Jan2026,A Pragmatist Robot: Learning to Plan Tasks by Experiencing the Real World,https://arxiv.org/abs/2507.16713,False,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy;execution_multimodal,8,long-term-memory
+2507.22925,2025-07-23,baseline-2025H2-Jan2026,Hierarchical Memory for High-Efficiency Long-Term Reasoning in LLM Agents,https://arxiv.org/abs/2507.22925,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,retrieval_representation,6,long-term-memory
+2507.21407,2025-07-29,baseline-2025H2-Jan2026,Graph-Augmented Large Language Model Agents: Current Progress and Future Prospects,https://arxiv.org/abs/2507.21407,False,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui,retrieval_representation,0,long-term-memory
+2508.00031,2025-07-30,baseline-2025H2-Jan2026,Git Context Controller: Manage the Context of LLM-based Agents like Git,https://arxiv.org/abs/2508.00031,False,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;systems_efficiency,,5,agent-memory
+2508.01287,2025-08-02,baseline-2025H2-Jan2026,Exploitation Is All You Need... for Exploration,https://arxiv.org/abs/2508.01287,False,peripheral,systems_cost,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis,,4,agent-memory
+2508.01415,2025-08-02,baseline-2025H2-Jan2026,RoboMemory: A Brain-inspired Multi-memory Agentic Framework for Interactive Environmental Learning in Physical Embodied Systems,https://arxiv.org/abs/2508.01415,True,core,update_forgetting,model,representation_organization;update_forgetting;execution_state;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;execution_multimodal,6,agent-memory
+2508.03341,2025-08-05,baseline-2025H2-Jan2026,What Deserves Memory: Adaptive Memory Distillation for LLM Agents,https://arxiv.org/abs/2508.03341,True,supporting,experience_learning,model,update_forgetting;experience_skill_learning;systems_efficiency,experience_skill_policy,3,episodic-memory
+2508.05002,2025-08-07,baseline-2025H2-Jan2026,AgenticData: An Agentic Data Analytics System for Heterogeneous Data,https://arxiv.org/abs/2508.05002,False,supporting,execution_state,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;multi_agent_shared,,5,agent-memory
+2508.07010,2025-08-09,baseline-2025H2-Jan2026,Narrative Memory in Machines: Multi-Agent Arc Extraction in Serialized TV,https://arxiv.org/abs/2508.07010,True,supporting,execution_state,model,representation_organization;update_forgetting;multi_agent_shared;multimodal_embodied_gui,personal_shared,4,episodic-memory
+2508.08997,2025-08-12,baseline-2025H2-Jan2026,Intrinsic Memory Agents: Heterogeneous Multi-Agent LLM Systems through Structured Contextual Memory,https://arxiv.org/abs/2508.08997,True,core,update_forgetting,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;systems_efficiency,retrieval_representation;personal_shared,3,agent-memory
+2508.09486,2025-08-13,baseline-2025H2-Jan2026,Video-EM: Event-Centric Episodic Memory for Long-Form Video Understanding,https://arxiv.org/abs/2508.09486,True,supporting,experience_learning,model,retrieval_access;experience_skill_learning;multimodal_embodied_gui,execution_multimodal,0,agent-memory;episodic-memory
+2508.12630,2025-08-18,baseline-2025H2-Jan2026,Semantic Anchoring in Agentic Memory: Leveraging Linguistic Structures for Persistent Conversational Context,https://arxiv.org/abs/2508.12630,True,supporting,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,retrieval_representation;execution_multimodal,7,agent-memory
+2508.13250,2025-08-18,baseline-2025H2-Jan2026,Explicit v.s. Implicit Memory: Exploring Multi-hop Complex Reasoning Over Personalized Information,https://arxiv.org/abs/2508.13250,True,core,update_forgetting,model,retrieval_access;evaluation_diagnosis,personal_shared,5,agent-memory
+2508.15294,2025-08-21,baseline-2025H2-Jan2026,A Multi-Memory Segment System for Generating High-Quality Long-Term Memory Content in Agents,https://arxiv.org/abs/2508.15294,True,core,update_forgetting,model,retrieval_access;evaluation_diagnosis,,3,agent-memory
+2508.16153,2025-08-22,baseline-2025H2-Jan2026,Memento: Fine-tuning LLM Agents without Fine-tuning LLMs,https://arxiv.org/abs/2508.16153,False,core,update_forgetting,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,3,episodic-memory
+2508.19005,2025-08-26,baseline-2025H2-Jan2026,Building Self-Evolving Agents via Experience-Driven Lifelong Learning: A Framework and Benchmark,https://arxiv.org/abs/2508.19005,False,supporting,execution_state,model,representation_organization;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis;experience_skill_policy,2,agent-memory
+2509.00997,2025-08-31,baseline-2025H2-Jan2026,Supporting Our AI Overlords: Redesigning Data Systems to be Agent-First,https://arxiv.org/abs/2509.00997,False,supporting,execution_state,model,retrieval_access;systems_efficiency,,0,agent-memory
+2509.11914,2025-09-15,baseline-2025H2-Jan2026,EgoMem: Lifelong Memory Agent for Full-duplex Omnimodal Models,https://arxiv.org/abs/2509.11914,True,core,update_forgetting,model,retrieval_access;update_forgetting;multimodal_embodied_gui,,6,agent-memory
+2509.21224,2025-09-25,baseline-2025H2-Jan2026,What Do LLM Agents Do When Left Alone? Evidence of Spontaneous Meta-Cognitive Patterns,https://arxiv.org/abs/2509.21224,False,core,execution_state,model,evaluation_diagnosis;multimodal_embodied_gui,,2,long-term-memory
+2509.25250,2025-09-27,baseline-2025H2-Jan2026,Memory Management and Contextual Consistency for Long-Running Low-Code Agents,https://arxiv.org/abs/2509.25250,True,core,update_forgetting,model,update_forgetting;multimodal_embodied_gui;systems_efficiency,,5,agent-memory
+2509.24704,2025-09-29,baseline-2025H2-Jan2026,MemGen: Weaving Generative Latent Memory for Self-Evolving Agents,https://arxiv.org/abs/2509.24704,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy,2,agent-memory
+2510.02373,2025-09-29,baseline-2025H2-Jan2026,A-MemGuard: A Proactive Defense Framework for LLM-Based Agent Memory,https://arxiv.org/abs/2510.02373,True,core,security_privacy,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,6,agent-memory
+2510.03612,2025-10-04,baseline-2025H2-Jan2026,Cross-Modal Content Optimization for Steering Web Agent Preferences,https://arxiv.org/abs/2510.03612,False,core,update_forgetting,model,evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,execution_multimodal,7,agent-memory
+2510.04618,2025-10-06,baseline-2025H2-Jan2026,Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models,https://arxiv.org/abs/2510.04618,False,core,update_forgetting,model,representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy,8,agent-memory
+2510.04851,2025-10-06,baseline-2025H2-Jan2026,LEGOMem: Modular Procedural Memory for Multi-agent LLM Systems for Workflow Automation,https://arxiv.org/abs/2510.04851,True,core,execution_state,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared,experience_skill_policy;personal_shared,2,agent-memory
+2510.05520,2025-10-07,baseline-2025H2-Jan2026,CAM: A Constructivist View of Agentic Memory for LLM-Based Reading Comprehension,https://arxiv.org/abs/2510.05520,True,core,access,model,retrieval_access;representation_organization;systems_efficiency,,2,agent-memory
+2510.08081,2025-10-09,baseline-2025H2-Jan2026,AutoQual: An LLM Agent for Automated Discovery of Interpretable Features for Review Quality Assessment,https://arxiv.org/abs/2510.08081,False,core,experience_learning,model,retrieval_access;experience_skill_learning,,3,long-term-memory
+2510.09720,2025-10-10,baseline-2025H2-Jan2026,Preference-Aware Memory Update for Long-Term LLM Agents,https://arxiv.org/abs/2510.09720,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;systems_efficiency,update_forgetting,5,long-term-memory
+2510.15966,2025-10-12,baseline-2025H2-Jan2026,PISA: A Pragmatic Psych-Inspired Unified Memory System for Enhanced AI Agency,https://arxiv.org/abs/2510.15966,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis,,4,agent-memory
+2510.11144,2025-10-13,baseline-2025H2-Jan2026,$How^{2}$: How to learn from procedural How-to questions,https://arxiv.org/abs/2510.11144,False,core,experience_learning,model,execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy,2,agent-memory
+2510.13896,2025-10-14,baseline-2025H2-Jan2026,"GenCellAgent: Generalizable, Training-Free Cellular Image Segmentation via Large Language Model Agents",https://arxiv.org/abs/2510.13896,False,core,shared_memory,model,representation_organization;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui,,5,long-term-memory
+2510.15620,2025-10-17,baseline-2025H2-Jan2026,On-device Semantic Selection Made Low Latency and Memory Efficient with Monolithic Forwarding,https://arxiv.org/abs/2510.15620,False,core,execution_state,model,retrieval_access;evaluation_diagnosis;systems_efficiency,systems_efficiency,6,agent-memory
+2510.18515,2025-10-21,baseline-2025H2-Jan2026,Socialized Learning and Emergent Behaviors in Multi-Agent Systems based on Multimodal Large Language Models,https://arxiv.org/abs/2510.18515,False,core,shared_memory,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;execution_multimodal;personal_shared,5,episodic-memory
+2510.19747,2025-10-22,baseline-2025H2-Jan2026,Review of Tools for Zero-Code LLM Based Application Development,https://arxiv.org/abs/2510.19747,False,core,shared_memory,model,multimodal_embodied_gui,,1,agent-memory
+2510.19897,2025-10-22,baseline-2025H2-Jan2026,Learning from Supervision with Semantic and Episodic Memory: A Reflective Approach to Agent Adaptation,https://arxiv.org/abs/2510.19897,True,core,representation,model,update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,4,episodic-memory
+2510.23730,2025-10-27,baseline-2025H2-Jan2026,Evaluating Long-Term Memory for Long-Context Question Answering,https://arxiv.org/abs/2510.23730,True,core,representation,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis,2,agent-memory;episodic-memory
+2510.25423,2025-10-29,baseline-2025H2-Jan2026,What Challenges Do Developers Face in AI Agent Systems? An Empirical Study on Stack Overflow & GitHub Issues,https://arxiv.org/abs/2510.25423,False,core,representation,model,retrieval_access;execution_state;evaluation_diagnosis,benchmark_diagnosis,2,agent-memory
+2510.26536,2025-10-30,baseline-2025H2-Jan2026,"RoboOS-NeXT: A Unified Memory-based Framework for Lifelong, Scalable, and Robust Multi-Robot Collaboration",https://arxiv.org/abs/2510.26536,True,core,representation,model,retrieval_access;representation_organization;execution_state;multi_agent_shared;multimodal_embodied_gui,execution_multimodal;systems_efficiency,4,agent-memory
+2510.27418,2025-10-31,baseline-2025H2-Jan2026,Dynamic Affective Memory Management for Personalized LLM Agents,https://arxiv.org/abs/2510.27418,True,core,representation,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis,personal_shared,3,long-term-memory
+2511.01448,2025-11-03,baseline-2025H2-Jan2026,LiCoMemory: Lightweight and Cognitive Agentic Memory for Efficient Long-Term Reasoning,https://arxiv.org/abs/2511.01448,True,core,representation,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,systems_efficiency,6,agent-memory;long-term-memory
+2511.02424,2025-11-04,baseline-2025H2-Jan2026,ReAcTree: Hierarchical LLM Agent Trees with Control Flow for Long-Horizon Task Planning,https://arxiv.org/abs/2511.02424,False,core,representation,model,retrieval_access;representation_organization;execution_state;multimodal_embodied_gui,retrieval_representation;execution_multimodal,9,episodic-memory
+2511.03475,2025-11-05,baseline-2025H2-Jan2026,ContextPilot: Fast Long-Context Inference via Context Reuse,https://arxiv.org/abs/2511.03475,False,core,access,model,retrieval_access;representation_organization;evaluation_diagnosis;multi_agent_shared;systems_efficiency,,5,agent-memory
+2511.06179,2025-11-09,baseline-2025H2-Jan2026,MemoriesDB: A Temporal-Semantic-Relational Database for Long-Term Agent Memory / Modeling Experience as a Graph of Temporal-Semantic Surfaces,https://arxiv.org/abs/2511.06179,True,core,experience_learning,model,retrieval_access;representation_organization;experience_skill_learning;systems_efficiency,retrieval_representation;experience_skill_policy,0,agent-memory
+2511.07587,2025-11-10,baseline-2025H2-Jan2026,Beyond Fact Retrieval: Episodic Memory for RAG with Generative Semantic Workspaces,https://arxiv.org/abs/2511.07587,True,core,representation,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,retrieval_representation,6,episodic-memory
+2511.08301,2025-11-11,baseline-2025H2-Jan2026,Smarter Together: Creating Agentic Communities of Practice through Shared Experiential Learning,https://arxiv.org/abs/2511.08301,False,core,experience_learning,model,experience_skill_learning;evaluation_diagnosis;multi_agent_shared,experience_skill_policy,5,agent-memory
+2511.12027,2025-11-15,baseline-2025H2-Jan2026,GCAgent: Long-Video Understanding via Schematic and Narrative Episodic Memory,https://arxiv.org/abs/2511.12027,True,core,execution_state,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,6,episodic-memory
+2511.12997,2025-11-17,baseline-2025H2-Jan2026,WebCoach: Self-Evolving Web Agents with Cross-Session Memory Guidance,https://arxiv.org/abs/2511.12997,True,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy;execution_multimodal,4,episodic-memory
+2511.16108,2025-11-20,baseline-2025H2-Jan2026,SkyRL-Agent: Efficient RL Training for Multi-turn LLM Agent,https://arxiv.org/abs/2511.16108,False,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;systems_efficiency,6,agent-memory
+2511.17775,2025-11-21,baseline-2025H2-Jan2026,Episodic Memory in Agentic Frameworks: Suggesting Next Tasks,https://arxiv.org/abs/2511.17775,True,core,update_forgetting,model,retrieval_access;multimodal_embodied_gui,,0,episodic-memory
+2511.21730,2025-11-21,baseline-2025H2-Jan2026,A Benchmark for Procedural Memory Retrieval in Language Agents,https://arxiv.org/abs/2511.21730,True,core,access,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis;retrieval_representation;experience_skill_policy,5,procedural-memory
+2511.18112,2025-11-22,baseline-2025H2-Jan2026,EchoVLA: Synergistic Declarative Memory for VLA-Driven Mobile Manipulation,https://arxiv.org/abs/2511.18112,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,,4,episodic-memory
+2511.18423,2025-11-23,baseline-2025H2-Jan2026,General Agentic Memory Via Deep Research,https://arxiv.org/abs/2511.18423,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;multimodal_embodied_gui,retrieval_representation,3,agent-memory
+2511.19192,2025-11-24,baseline-2025H2-Jan2026,AME: An Efficient Heterogeneous Agentic Memory Engine for Smartphones,https://arxiv.org/abs/2511.19192,True,core,update_forgetting,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,systems_efficiency,4,agent-memory
+2511.20297,2025-11-25,baseline-2025H2-Jan2026,Improving Language Agents through BREW,https://arxiv.org/abs/2511.20297,False,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;systems_efficiency,,2,agent-memory
+2511.20857,2025-11-25,baseline-2025H2-Jan2026,Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory,https://arxiv.org/abs/2511.20857,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;experience_skill_policy,5,memory-evaluation
+2512.00742,2025-11-30,baseline-2025H2-Jan2026,On the Regulatory Potential of User Interfaces for AI Agent Governance,https://arxiv.org/abs/2512.00742,False,core,access,model,representation_organization;security_privacy_trust,security_governance,0,agent-memory
+2512.02227,2025-12-01,baseline-2025H2-Jan2026,Orchestration Framework for Financial Agents: From Algorithmic Trading to Agentic Trading,https://arxiv.org/abs/2512.02227,False,core,access,model,representation_organization,,1,agent-memory
+2512.02228,2025-12-01,baseline-2025H2-Jan2026,"STRIDE: A Systematic Framework for Selecting AI Modalities -- Agentic AI, AI Assistants, or LLM Calls",https://arxiv.org/abs/2512.02228,False,core,access,model,representation_organization;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,4,long-term-memory
+2512.02425,2025-12-02,baseline-2025H2-Jan2026,WorldMM: Dynamic Multimodal Memory Agent for Long Video Reasoning,https://arxiv.org/abs/2512.02425,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,8,agent-memory;episodic-memory
+2512.02458,2025-12-02,baseline-2025H2-Jan2026,Vision to Geometry: 3D Spatial Memory for Sequential Embodied MLLM Reasoning and Exploration,https://arxiv.org/abs/2512.02458,True,core,update_forgetting,model,retrieval_access;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,3,agent-memory
+2512.03627,2025-12-03,baseline-2025H2-Jan2026,MemVerse: Multimodal Memory for Lifelong Learning Agents,https://arxiv.org/abs/2512.03627,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;execution_multimodal,4,agent-memory
+2512.04668,2025-12-04,baseline-2025H2-Jan2026,Topology Matters: Measuring Memory Leakage in Multi-Agent LLMs,https://arxiv.org/abs/2512.04668,True,core,execution_state,model,representation_organization;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui,security_governance;personal_shared,4,agent-memory
+2512.06688,2025-12-07,baseline-2025H2-Jan2026,PersonaMem-v2: Towards Personalized Intelligence via Learning Implicit User Personas and Agentic Memory,https://arxiv.org/abs/2512.06688,True,core,application,model,experience_skill_learning;systems_efficiency,experience_skill_policy;personal_shared,6,agent-memory
+2512.09458,2025-12-10,baseline-2025H2-Jan2026,Architectures for Building Agentic AI,https://arxiv.org/abs/2512.09458,False,core,execution_state,model,execution_state;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,,1,agent-memory
+2512.10166,2025-12-10,baseline-2025H2-Jan2026,Emergent Collective Memory in Decentralized Multi-Agent AI Systems,https://arxiv.org/abs/2512.10166,True,core,update_forgetting,model,representation_organization;evaluation_diagnosis;multi_agent_shared,personal_shared,7,agent-memory
+2512.10696,2025-12-11,baseline-2025H2-Jan2026,"Remember Me, Refine Me: A Dynamic Procedural Memory Framework for Experience-Driven Agent Evolution",https://arxiv.org/abs/2512.10696,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;systems_efficiency,experience_skill_policy,7,agent-memory;procedural-memory
+2512.11303,2025-12-12,baseline-2025H2-Jan2026,Unifying Dynamic Tool Creation and Cross-Task Experience Sharing through Cognitive Memory Architecture,https://arxiv.org/abs/2512.11303,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;systems_efficiency,experience_skill_policy,8,agent-memory;episodic-memory
+2512.12686,2025-12-14,baseline-2025H2-Jan2026,Memoria: A Scalable Agentic Memory Framework for Personalized Conversational AI,https://arxiv.org/abs/2512.12686,True,core,access,model,representation_organization;experience_skill_learning;systems_efficiency,systems_efficiency;personal_shared,0,agent-memory
+2512.12818,2025-12-14,baseline-2025H2-Jan2026,"Hindsight is 20/20: Building Agent Memory that Retains, Recalls, and Reflects",https://arxiv.org/abs/2512.12818,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,retrieval_representation,7,agent-memory;memory-evaluation
+2512.12856,2025-12-14,baseline-2025H2-Jan2026,Forgetful but Faithful: A Cognitive Memory Architecture and Benchmark for Privacy-Aware Generative Agents,https://arxiv.org/abs/2512.12856,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;update_forgetting;security_governance,2,agent-memory
+2512.12967,2025-12-15,baseline-2025H2-Jan2026,QwenLong-L1.5: Post-Training Recipe for Long-Context Reasoning and Memory Management,https://arxiv.org/abs/2512.12967,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;systems_efficiency,,7,agent-memory
+2512.13564,2025-12-15,baseline-2025H2-Jan2026,Memory in the Age of AI Agents,https://arxiv.org/abs/2512.13564,True,core,access,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,,3,agent-memory;memory-evaluation
+2512.16301,2025-12-18,baseline-2025H2-Jan2026,"Adaptation of Agentic AI: A Survey of Post-Training, Memory, and Skills",https://arxiv.org/abs/2512.16301,True,core,experience_learning,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;experience_skill_policy,4,long-term-memory
+2512.16962,2025-12-18,baseline-2025H2-Jan2026,MemoryGraft: Persistent Compromise of LLM Agents via Poisoned Experience Retrieval,https://arxiv.org/abs/2512.16962,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;security_privacy_trust,retrieval_representation;experience_skill_policy;security_governance,3,agent-memory;long-term-memory
+2512.18337,2025-12-20,baseline-2025H2-Jan2026,Towards Efficient Agents: A Co-Design of Inference Architecture and System,https://arxiv.org/abs/2512.18337,False,core,execution_state,model,representation_organization;execution_state;evaluation_diagnosis;systems_efficiency,systems_efficiency,4,agent-memory
+2512.18571,2025-12-21,baseline-2025H2-Jan2026,ESearch-R1: Learning Cost-Aware MLLM Agents for Interactive Embodied Search via Reinforcement Learning,https://arxiv.org/abs/2512.18571,False,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,retrieval_representation;experience_skill_policy;execution_multimodal;systems_efficiency,6,episodic-memory
+2512.18746,2025-12-21,baseline-2025H2-Jan2026,MemEvolve: Meta-Evolution of Agent Memory Systems,https://arxiv.org/abs/2512.18746,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis,experience_skill_policy,2,agent-memory
+2512.18950,2025-12-22,baseline-2025H2-Jan2026,Learning Hierarchical Procedural Memory for LLM Agents through Bayesian Selection and Contrastive Refinement,https://arxiv.org/abs/2512.18950,True,core,update_forgetting,model,representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;systems_efficiency,retrieval_representation;experience_skill_policy,5,procedural-memory
+2512.19537,2025-12-22,baseline-2025H2-Jan2026,Event Extraction in Large Language Model,https://arxiv.org/abs/2512.19537,False,core,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui,,3,agent-memory
+2512.20111,2025-12-23,baseline-2025H2-Jan2026,ABBEL: Learning Natural-Language Belief States for Memory-Efficient Interaction,https://arxiv.org/abs/2512.20111,False,core,update_forgetting,model,update_forgetting;execution_state;experience_skill_learning;systems_efficiency,experience_skill_policy;systems_efficiency,7,agent-memory
+2512.21567,2025-12-25,baseline-2025H2-Jan2026,Beyond Heuristics: A Decision-Theoretic Framework for Agent Memory Management,https://arxiv.org/abs/2512.21567,True,core,update_forgetting,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,,3,agent-memory
+2512.22716,2025-12-27,baseline-2025H2-Jan2026,Memento 2: Learning by Stateful Reflective Memory,https://arxiv.org/abs/2512.22716,True,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,2,episodic-memory
+2512.23343,2025-12-29,baseline-2025H2-Jan2026,AI Meets Brain: Memory Systems from Cognitive Neuroscience to Autonomous Agents,https://arxiv.org/abs/2512.23343,True,core,representation,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,,2,agent-memory
+2601.01885,2026-01-05,baseline-2025H2-Jan2026,Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents,https://arxiv.org/abs/2601.01885,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy,5,agent-memory;long-term-memory
+2601.02577,2026-01-05,baseline-2025H2-Jan2026,Orchestral AI: A Framework for Agent Orchestration,https://arxiv.org/abs/2601.02577,False,core,representation,model,representation_organization;systems_efficiency,,0,agent-memory
+2601.02732,2026-01-06,baseline-2025H2-Jan2026,Agentic Memory Enhanced Recursive Reasoning for Root Cause Localization in Microservices,https://arxiv.org/abs/2601.02732,True,core,update_forgetting,model,multi_agent_shared;systems_efficiency,,5,agent-memory
+2601.02744,2026-01-06,baseline-2025H2-Jan2026,SYNAPSE: Empowering LLM Agents with Episodic-Semantic Memory via Spreading Activation,https://arxiv.org/abs/2601.02744,True,core,update_forgetting,model,retrieval_access;evaluation_diagnosis,,6,agent-memory
+2601.03236,2026-01-06,baseline-2025H2-Jan2026,MAGMA: A Multi-Graph based Agentic Memory Architecture for AI Agents,https://arxiv.org/abs/2601.03236,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;multimodal_embodied_gui,retrieval_representation,4,agent-memory
+2601.03543,2026-01-07,baseline-2025H2-Jan2026,EvolMem: A Cognitive-Driven Benchmark for Multi-Session Dialogue Memory,https://arxiv.org/abs/2601.03543,True,core,representation,model,evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis,6,agent-memory
+2601.03785,2026-01-07,baseline-2025H2-Jan2026,Membox: Weaving Topic Continuity into Long-Range Memory for LLM Agents,https://arxiv.org/abs/2601.03785,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting,,4,agent-memory
+2601.04170,2026-01-07,baseline-2025H2-Jan2026,Agent Drift: Quantifying Behavioral Degradation in Multi-Agent LLM Systems Over Extended Interactions,https://arxiv.org/abs/2601.04170,False,supporting,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multi_agent_shared;systems_efficiency,personal_shared,1,episodic-memory
+2601.05504,2026-01-09,baseline-2025H2-Jan2026,Memory Poisoning Attack and Defense on Memory Based LLM-Agents,https://arxiv.org/abs/2601.05504,True,supporting,security_privacy,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,security_governance,6,long-term-memory
+2601.06282,2026-01-09,baseline-2025H2-Jan2026,Amory: Building Coherent Narrative-Driven Agent Memory through Agentic Reasoning,https://arxiv.org/abs/2601.06282,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,,8,agent-memory
+2601.06377,2026-01-10,baseline-2025H2-Jan2026,HiMem: Hierarchical Long-Term Memory for LLM Long-Horizon Agents,https://arxiv.org/abs/2601.06377,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,retrieval_representation;execution_multimodal,6,episodic-memory
+2601.06411,2026-01-10,baseline-2025H2-Jan2026,Structured Episodic Event Memory,https://arxiv.org/abs/2601.06411,True,core,experience_learning,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust,retrieval_representation,6,episodic-memory
+2601.10744,2026-01-11,baseline-2025H2-Jan2026,Explore with Long-term Memory: A Benchmark and Multimodal LLM-based Reinforcement Learning Framework for Embodied Exploration,https://arxiv.org/abs/2601.10744,True,supporting,execution_state,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;experience_skill_policy;execution_multimodal,4,agent-memory;episodic-memory
+2601.07470,2026-01-12,baseline-2025H2-Jan2026,Learning How to Remember: A Meta-Cognitive Management Method for Structured and Transferable Agent Memory,https://arxiv.org/abs/2601.07470,True,core,experience_learning,model,representation_organization;execution_state;experience_skill_learning,retrieval_representation;experience_skill_policy,4,agent-memory
+2601.07779,2026-01-12,baseline-2025H2-Jan2026,OS-Symphony: A Holistic Framework for Robust and Generalist Computer-Using Agent,https://arxiv.org/abs/2601.07779,False,supporting,execution_state,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,,6,agent-memory
+2601.08160,2026-01-13,baseline-2025H2-Jan2026,SwiftMem: Fast Agentic Memory via Query-aware Indexing,https://arxiv.org/abs/2601.08160,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,retrieval_representation,5,agent-memory
+2601.08323,2026-01-13,baseline-2025H2-Jan2026,AtomMem : Learnable Dynamic Agentic Memory with Atomic Memory Operation,https://arxiv.org/abs/2601.08323,True,core,experience_learning,model,representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis,experience_skill_policy,4,agent-memory
+2601.09113,2026-01-14,baseline-2025H2-Jan2026,The AI Hippocampus: How Far are We From Human Memory?,https://arxiv.org/abs/2601.09113,True,supporting,experience_learning,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis,2,agent-memory
+2601.09636,2026-01-14,baseline-2025H2-Jan2026,PersonalAlign: Hierarchical Implicit Intent Alignment for Personalized GUI Agent with Long-Term User-Centric Records,https://arxiv.org/abs/2601.09636,False,core,execution_state,model,representation_organization;evaluation_diagnosis;multimodal_embodied_gui,retrieval_representation;execution_multimodal;personal_shared,5,agent-memory
+2602.06051,2026-01-14,baseline-2025H2-Jan2026,CAST: Character-and-Scene Episodic Memory for Agents,https://arxiv.org/abs/2602.06051,True,core,representation,model,retrieval_access;representation_organization;experience_skill_learning,,6,agent-memory;episodic-memory
+2602.06052,2026-01-14,baseline-2025H2-Jan2026,Rethinking Memory Mechanisms of Foundation Agents in the Second Half: A Survey,https://arxiv.org/abs/2602.06052,True,core,access,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis,2,agent-memory
+2601.10702,2026-01-15,baseline-2025H2-Jan2026,Grounding Agent Memory in Contextual Intent,https://arxiv.org/abs/2601.10702,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis,,6,agent-memory
+2601.11854,2026-01-17,baseline-2025H2-Jan2026,ATOD: An Evaluation Framework and Benchmark for Agentic Task-Oriented Dialogue Systems,https://arxiv.org/abs/2601.11854,False,core,representation,model,execution_state;evaluation_diagnosis,benchmark_diagnosis,4,agent-memory
+2601.12771,2026-01-19,baseline-2025H2-Jan2026,Who Does This Name Remind You of ? Nationality Prediction via Large Language Model Associative Memory,https://arxiv.org/abs/2601.12771,True,core,update_forgetting,model,retrieval_access;multi_agent_shared;multimodal_embodied_gui,,5,agent-memory
+2601.14192,2026-01-20,baseline-2025H2-Jan2026,"Toward Efficient Agents: Memory, Tool learning, and Planning",https://arxiv.org/abs/2601.14192,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy;systems_efficiency,3,agent-memory
+2601.14735,2026-01-21,baseline-2025H2-Jan2026,Optimizing FaaS Platforms for MCP-enabled Agentic Workflows,https://arxiv.org/abs/2601.14735,False,core,execution_state,model,retrieval_access;evaluation_diagnosis;multi_agent_shared;systems_efficiency,,4,agent-memory
+2601.16690,2026-01-23,baseline-2025H2-Jan2026,EMemBench: Interactive Benchmarking of Episodic Memory for VLM Agents,https://arxiv.org/abs/2601.16690,True,core,experience_learning,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,benchmark_diagnosis,6,agent-memory
+2601.16872,2026-01-23,baseline-2025H2-Jan2026,From Atom to Community: Structured and Evolving Agent Memory for User Behavior Modeling,https://arxiv.org/abs/2601.16872,True,core,update_forgetting,model,representation_organization;update_forgetting,retrieval_representation;experience_skill_policy,4,agent-memory
+2601.17887,2026-01-25,baseline-2025H2-Jan2026,When Personalization Legitimizes Risks: Uncovering Safety Vulnerabilities in Personalized Dialogue Agents,https://arxiv.org/abs/2601.17887,False,core,execution_state,model,representation_organization;experience_skill_learning;evaluation_diagnosis;security_privacy_trust,security_governance;personal_shared,6,long-term-memory
+2601.18642,2026-01-26,baseline-2025H2-Jan2026,FadeMem: Biologically-Inspired Forgetting for Efficient Agent Memory,https://arxiv.org/abs/2601.18642,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,update_forgetting;systems_efficiency,3,agent-memory
+2601.20352,2026-01-28,baseline-2025H2-Jan2026,AMA: Adaptive Memory via Multi-Agent Collaboration,https://arxiv.org/abs/2601.20352,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multi_agent_shared;systems_efficiency,experience_skill_policy;personal_shared,7,agent-memory;long-term-memory
+2601.20465,2026-01-28,baseline-2025H2-Jan2026,BMAM: Brain-inspired Multi-Agent Memory Framework,https://arxiv.org/abs/2601.20465,True,core,execution_state,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;multi_agent_shared;systems_efficiency,personal_shared,4,agent-memory
+2601.21468,2026-01-29,baseline-2025H2-Jan2026,MemOCR: Layout-Aware Visual Memory for Efficient Long-Horizon Reasoning,https://arxiv.org/abs/2601.21468,True,core,update_forgetting,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal;systems_efficiency,4,agent-memory
+2601.21714,2026-01-29,baseline-2025H2-Jan2026,E-mem: Multi-agent based Episodic Context Reconstruction for LLM Agent Memory,https://arxiv.org/abs/2601.21714,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;systems_efficiency,personal_shared,2,agent-memory
+2601.21841,2026-01-29,baseline-2025H2-Jan2026,Embodied Task Planning via Graph-Informed Action Generation with Large Language Models,https://arxiv.org/abs/2601.21841,False,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,retrieval_representation;execution_multimodal,7,agent-memory
+2601.22964,2026-01-30,baseline-2025H2-Jan2026,EvoClinician: A Self-Evolving Agent for Multi-Turn Medical Diagnosis via Test-Time Evolutionary Learning,https://arxiv.org/abs/2601.22964,False,core,update_forgetting,model,update_forgetting;experience_skill_learning;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;experience_skill_policy,5,agent-memory
+2601.22974,2026-01-30,baseline-2025H2-Jan2026,MiTa: A Hierarchical Multi-Agent Collaboration Framework with Memory-integrated and Task Allocation,https://arxiv.org/abs/2601.22974,True,core,update_forgetting,model,representation_organization;update_forgetting;execution_state;multi_agent_shared;multimodal_embodied_gui,retrieval_representation;personal_shared,4,episodic-memory
+2601.23014,2026-01-30,baseline-2025H2-Jan2026,Mem-T: Densifying Rewards for Long-Horizon Memory Agents,https://arxiv.org/abs/2601.23014,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,execution_multimodal,3,agent-memory
+2602.00352,2026-01-30,baseline-2025H2-Jan2026,DETOUR: An Interactive Benchmark for Dual-Agent Search and Reasoning,https://arxiv.org/abs/2602.00352,False,core,access,model,retrieval_access;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;retrieval_representation,4,agent-memory
+2602.00364,2026-01-30,baseline-2025H2-Jan2026,"""Someone Hid It"": Query-Agnostic Black-Box Attacks on LLM-Based Retrieval",https://arxiv.org/abs/2602.00364,False,core,access,model,retrieval_access;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation;security_governance,2,agent-memory
+2602.00675,2026-01-31,baseline-2025H2-Jan2026,Factored Reasoning with Inner Speech and Persistent Memory for Evidence-Grounded Human-Robot Interaction,https://arxiv.org/abs/2602.00675,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,3,agent-memory
+2602.01869,2026-02-02,transition-2026-Feb-Mar,Skill-Pro: Learning Reusable Skills from Experience via Non-Parametric PPO for LLM Agents,https://arxiv.org/abs/2602.01869,False,core,update_forgetting,model,update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,3,procedural-memory
+2602.02007,2026-02-02,transition-2026-Feb-Mar,Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation,https://arxiv.org/abs/2602.02007,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;multimodal_embodied_gui;systems_efficiency,retrieval_representation,2,agent-memory
+2602.02164,2026-02-02,transition-2026-Feb-Mar,Co-RedTeam: Orchestrated Security Discovery and Exploitation with LLM Agents,https://arxiv.org/abs/2602.02164,False,supporting,security_privacy,model,representation_organization;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared,security_governance,7,long-term-memory
+2602.02369,2026-02-02,transition-2026-Feb-Mar,Live-Evo: Online Evolution of Agentic Memory from Continuous Feedback,https://arxiv.org/abs/2602.02369,True,core,update_forgetting,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,,5,agent-memory
+2602.02474,2026-02-02,transition-2026-Feb-Mar,MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents,https://arxiv.org/abs/2602.02474,True,core,update_forgetting,model,representation_organization;update_forgetting;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,5,agent-memory
+2602.03036,2026-02-03,transition-2026-Feb-Mar,LatentMem: Customizing Latent Memory for Multi-Agent Systems,https://arxiv.org/abs/2602.03036,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;systems_efficiency,personal_shared,5,agent-memory
+2602.03224,2026-02-03,transition-2026-Feb-Mar,TAME: A Trustworthy Test-Time Evolution of Agent Memory with Systematic Benchmarking,https://arxiv.org/abs/2602.03224,True,core,security_privacy,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui,benchmark_diagnosis;security_governance,2,agent-memory
+2602.03315,2026-02-03,transition-2026-Feb-Mar,Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity,https://arxiv.org/abs/2602.03315,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,retrieval_representation,2,agent-memory
+2602.04482,2026-02-04,transition-2026-Feb-Mar,ProAgentBench: Evaluating LLM Agents for Proactive Assistance with Real-World Data,https://arxiv.org/abs/2602.04482,False,supporting,application,model,representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,benchmark_diagnosis,7,long-term-memory
+2602.04640,2026-02-04,transition-2026-Feb-Mar,"Towards Structured, State-Aware, and Execution-Grounded Reasoning for Software Engineering Agents",https://arxiv.org/abs/2602.04640,False,core,update_forgetting,model,representation_organization;execution_state,retrieval_representation,0,agent-memory
+2603.00026,2026-02-04,transition-2026-Feb-Mar,ActMem: Bridging the Gap Between Memory Retrieval and Reasoning in LLM Agents,https://arxiv.org/abs/2603.00026,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis,retrieval_representation,6,memory-evaluation
+2602.05665,2026-02-05,transition-2026-Feb-Mar,"Graph-based Agent Memory: Taxonomy, Techniques, and Applications",https://arxiv.org/abs/2602.05665,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;retrieval_representation,5,agent-memory
+2602.06025,2026-02-05,transition-2026-Feb-Mar,Learning Query-Aware Budget-Tier Routing for Runtime Agent Memory,https://arxiv.org/abs/2602.06025,True,supporting,application,model,retrieval_access;representation_organization;experience_skill_learning;systems_efficiency,experience_skill_policy;systems_efficiency,2,agent-memory
+2602.17692,2026-02-06,transition-2026-Feb-Mar,Agentic Unlearning: When LLM Agent Meets Machine Unlearning,https://arxiv.org/abs/2602.17692,False,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy;security_governance,2,long-term-memory
+2603.03296,2026-02-06,transition-2026-Feb-Mar,PlugMem: A Task-Agnostic Plugin Memory Module for LLM Agents,https://arxiv.org/abs/2603.03296,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,5,episodic-memory;long-term-memory
+2602.07398,2026-02-07,transition-2026-Feb-Mar,AgentSys: Secure and Dynamic LLM Agents Through Explicit Hierarchical Memory Management,https://arxiv.org/abs/2602.07398,True,core,update_forgetting,model,representation_organization;evaluation_diagnosis;security_privacy_trust,retrieval_representation,6,agent-memory
+2602.07624,2026-02-07,transition-2026-Feb-Mar,M2A: Multimodal Memory Agent with Dual-Layer Hybrid Memory for Long-Term Personalized Interactions,https://arxiv.org/abs/2602.07624,True,core,update_forgetting,model,retrieval_access;update_forgetting;multimodal_embodied_gui;systems_efficiency,execution_multimodal;personal_shared,6,agent-memory
+2602.07755,2026-02-08,transition-2026-Feb-Mar,Learning to Continually Learn via Meta-learning Agentic Memory Designs,https://arxiv.org/abs/2602.07755,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy,3,agent-memory
+2602.07885,2026-02-08,transition-2026-Feb-Mar,MemFly: On-the-Fly Memory Optimization via Information Bottleneck,https://arxiv.org/abs/2602.07885,True,core,update_forgetting,model,retrieval_access;representation_organization;systems_efficiency,,4,long-term-memory
+2602.08369,2026-02-09,transition-2026-Feb-Mar,MemAdapter: Fast Alignment across Agent Memory Paradigms via Generative Subgraph Retrieval,https://arxiv.org/abs/2602.08369,True,core,update_forgetting,model,retrieval_access;execution_state;evaluation_diagnosis;systems_efficiency,retrieval_representation;experience_skill_policy,6,agent-memory
+2602.09319,2026-02-10,transition-2026-Feb-Mar,Benchmarking Knowledge-Extraction Attack and Defense on Retrieval-Augmented Generation,https://arxiv.org/abs/2602.09319,False,core,access,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,benchmark_diagnosis;retrieval_representation;security_governance;systems_efficiency,2,agent-memory
+2602.09712,2026-02-10,transition-2026-Feb-Mar,TraceMem: Weaving Narrative Memory Schemata from User Conversational Traces,https://arxiv.org/abs/2602.09712,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis,,6,episodic-memory
+2602.10715,2026-02-11,transition-2026-Feb-Mar,Locomo-Plus: Beyond-Factual Cognitive Memory Evaluation Framework for LLM Agents,https://arxiv.org/abs/2602.10715,True,core,execution_state,model,retrieval_access;evaluation_diagnosis,benchmark_diagnosis,4,memory-evaluation
+2602.11243,2026-02-11,transition-2026-Feb-Mar,Evaluating Memory Structure in LLM Agents,https://arxiv.org/abs/2602.11243,True,core,representation,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;retrieval_representation,3,agent-memory;long-term-memory;memory-evaluation
+2602.13530,2026-02-13,transition-2026-Feb-Mar,REMem: Reasoning with Episodic Memory in Language Agent,https://arxiv.org/abs/2602.13530,True,core,representation,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis,,6,memory-evaluation
+2602.18493,2026-02-13,transition-2026-Feb-Mar,Learning to Remember: End-to-End Training of Memory Agents for Long-Context Reasoning,https://arxiv.org/abs/2602.18493,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis,experience_skill_policy,4,agent-memory
+2602.13594,2026-02-14,transition-2026-Feb-Mar,Hippocampus: An Efficient and Scalable Memory Module for Agentic AI,https://arxiv.org/abs/2602.13594,True,core,execution_state,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;systems_efficiency,systems_efficiency,2,agent-memory
+2602.13933,2026-02-15,transition-2026-Feb-Mar,HyMem: Hybrid Memory Architecture with Dynamic Retrieval Scheduling,https://arxiv.org/abs/2602.13933,True,core,execution_state,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;systems_efficiency,retrieval_representation,6,long-term-memory
+2602.14038,2026-02-15,transition-2026-Feb-Mar,Choosing How to Remember: Adaptive Memory Structures for LLM Agents,https://arxiv.org/abs/2602.14038,True,core,execution_state,model,representation_organization;execution_state;evaluation_diagnosis,retrieval_representation;experience_skill_policy,4,agent-memory
+2602.15274,2026-02-17,transition-2026-Feb-Mar,When Remembering and Planning are Worth it: Navigating under Change,https://arxiv.org/abs/2602.15274,True,core,execution_state,model,retrieval_access;execution_state;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,,0,agent-memory
+2602.15513,2026-02-17,transition-2026-Feb-Mar,HIMM: Human-Inspired Long-Term Memory Modeling for Embodied Exploration and Question Answering,https://arxiv.org/abs/2602.15513,True,core,execution_state,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,4,episodic-memory
+2602.15654,2026-02-17,transition-2026-Feb-Mar,Zombie Agents: Persistent Control of Self-Evolving LLM Agents via Self-Reinforcing Injections,https://arxiv.org/abs/2602.15654,False,core,execution_state,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,experience_skill_policy,3,long-term-memory
+2603.02240,2026-02-17,transition-2026-Feb-Mar,SuperLocalMemory: Privacy-Preserving Multi-Agent Memory with Bayesian Trust Defense Against Memory Poisoning,https://arxiv.org/abs/2603.02240,True,core,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;systems_efficiency,security_governance;systems_efficiency;personal_shared,5,agent-memory
+2603.04428,2026-02-17,transition-2026-Feb-Mar,Agent Memory Below the Prompt: Persistent Q4 KV Cache for Multi-Agent LLM Inference on Edge Devices,https://arxiv.org/abs/2603.04428,True,core,execution_state,model,evaluation_diagnosis;multi_agent_shared;systems_efficiency,systems_efficiency;personal_shared,5,agent-memory
+2602.16313,2026-02-18,transition-2026-Feb-Mar,MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks,https://arxiv.org/abs/2602.16313,True,core,execution_state,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis,3,agent-memory;memory-evaluation
+2602.16493,2026-02-18,transition-2026-Feb-Mar,MMA: Multimodal Memory Agent,https://arxiv.org/abs/2602.16493,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,8,agent-memory
+2602.19320,2026-02-22,transition-2026-Feb-Mar,Anatomy of Agentic Memory: Taxonomy and Empirical Analysis of Evaluation and System Limitations,https://arxiv.org/abs/2602.19320,True,core,execution_state,model,representation_organization;execution_state;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis,4,agent-memory
+2603.04443,2026-02-22,transition-2026-Feb-Mar,AMV-L: Lifecycle-Managed Agent Memory for Tail-Latency Control in Long-Running LLM Systems,https://arxiv.org/abs/2603.04443,True,core,execution_state,model,retrieval_access;update_forgetting;evaluation_diagnosis;systems_efficiency,update_forgetting;systems_efficiency,7,agent-memory;long-term-memory
+2602.21394,2026-02-24,transition-2026-Feb-Mar,MemoPhishAgent: Memory-Augmented Multi-Modal LLM Agent for Phishing URL Detection,https://arxiv.org/abs/2602.21394,True,core,execution_state,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,,8,episodic-memory
+2602.21477,2026-02-25,transition-2026-Feb-Mar,Pancake: Hierarchical Memory System for Multi-Agent LLM Serving,https://arxiv.org/abs/2602.21477,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;multi_agent_shared;systems_efficiency,retrieval_representation;systems_efficiency;personal_shared,6,agent-memory
+2602.22406,2026-02-25,transition-2026-Feb-Mar,Towards Autonomous Memory Agents,https://arxiv.org/abs/2602.22406,True,core,update_forgetting,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;systems_efficiency,,6,agent-memory
+2602.22769,2026-02-26,transition-2026-Feb-Mar,AMA-Bench: Evaluating Long-Horizon Memory for Agentic Applications,https://arxiv.org/abs/2602.22769,True,core,update_forgetting,model,retrieval_access;execution_state;evaluation_diagnosis,benchmark_diagnosis;execution_multimodal,8,agent-memory;memory-evaluation
+2602.23720,2026-02-27,transition-2026-Feb-Mar,The Auton Agentic AI Framework,https://arxiv.org/abs/2602.23720,False,core,execution_state,model,representation_organization;update_forgetting;experience_skill_learning;security_privacy_trust;systems_efficiency,,0,episodic-memory
+2602.23937,2026-02-27,transition-2026-Feb-Mar,Enhancing Vision-Language Navigation with Multimodal Event Knowledge from Real-World Indoor Tour Videos,https://arxiv.org/abs/2602.23937,False,core,execution_state,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal;systems_efficiency,4,episodic-memory
+2603.01160,2026-03-01,transition-2026-Feb-Mar,Semantic XPath: Structured Agentic Memory Access for Conversational AI,https://arxiv.org/abs/2603.01160,True,core,access,model,retrieval_access;representation_organization;update_forgetting;multimodal_embodied_gui;systems_efficiency,retrieval_representation,4,agent-memory
+2603.01966,2026-03-02,transition-2026-Feb-Mar,AMemGym: Interactive Memory Benchmarking for Assistants in Long-Horizon Conversations,https://arxiv.org/abs/2603.01966,True,core,update_forgetting,model,representation_organization;execution_state;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;execution_multimodal,2,agent-memory;memory-evaluation
+2603.02206,2026-03-02,transition-2026-Feb-Mar,VoiceAgentRAG: Solving the RAG Latency Bottleneck in Real-Time Voice Agents Using Dual-Agent Architectures,https://arxiv.org/abs/2603.02206,False,core,access,model,retrieval_access;representation_organization;systems_efficiency,retrieval_representation;systems_efficiency,1,agent-memory
+2603.02473,2026-03-02,transition-2026-Feb-Mar,Diagnosing Retrieval vs. Utilization Bottlenecks in LLM Agent Memory,https://arxiv.org/abs/2603.02473,True,core,access,model,retrieval_access;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;retrieval_representation,6,agent-memory
+2603.03024,2026-03-03,transition-2026-Feb-Mar,MA-CoNav: A Master-Slave Multi-Agent Framework with Hierarchical Collaboration and Dual-Level Reflection for Long-Horizon Embodied VLN,https://arxiv.org/abs/2603.03024,False,core,update_forgetting,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,retrieval_representation;execution_multimodal;personal_shared,4,agent-memory
+2603.03148,2026-03-03,transition-2026-Feb-Mar,From Language to Action: Can LLM-Based Agents Be Used for Embodied Robot Cognition?,https://arxiv.org/abs/2603.03148,False,core,execution_state,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,4,episodic-memory
+2603.15642,2026-03-03,transition-2026-Feb-Mar,CraniMem: Cranial Inspired Gated and Bounded Memory for Agentic Systems,https://arxiv.org/abs/2603.15642,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis,,5,agent-memory
+2603.04549,2026-03-04,transition-2026-Feb-Mar,Adaptive Memory Admission Control for LLM Agents,https://arxiv.org/abs/2603.04549,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,experience_skill_policy,7,long-term-memory
+2603.04740,2026-03-05,transition-2026-Feb-Mar,Memory as Ontology: A Constitutional Memory Architecture for Persistent Digital Citizens,https://arxiv.org/abs/2603.04740,True,core,execution_state,model,retrieval_access;representation_organization;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,retrieval_representation,0,agent-memory
+2603.23516,2026-03-06,transition-2026-Feb-Mar,MSA: Memory Sparse Attention for Efficient End-to-End Memory Model Scaling to 100M Tokens,https://arxiv.org/abs/2603.23516,True,core,update_forgetting,model,evaluation_diagnosis;systems_efficiency,systems_efficiency,5,agent-memory
+2603.08755,2026-03-07,transition-2026-Feb-Mar,Turn: A Language for Agentic Computation,https://arxiv.org/abs/2603.08755,False,core,update_forgetting,model,evaluation_diagnosis,,3,agent-memory
+2603.07392,2026-03-08,transition-2026-Feb-Mar,Can Large Language Models Keep Up? Benchmarking Online Adaptation to Continual Knowledge Streams,https://arxiv.org/abs/2603.07392,False,core,update_forgetting,model,representation_organization;evaluation_diagnosis,benchmark_diagnosis;experience_skill_policy;systems_efficiency,4,agent-memory
+2603.07670,2026-03-08,transition-2026-Feb-Mar,"Memory for Autonomous LLM Agents:Mechanisms, Evaluation, and Emerging Frontiers",https://arxiv.org/abs/2603.07670,True,core,representation,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis,2,agent-memory
+2603.07784,2026-03-08,transition-2026-Feb-Mar,ProgAgent:A Continual RL Agent with Progress-Aware Rewards,https://arxiv.org/abs/2603.07784,False,core,update_forgetting,model,update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,4,agent-memory
+2603.10062,2026-03-09,transition-2026-Feb-Mar,Multi-Agent Memory from a Computer Architecture Perspective: Visions and Challenges Ahead,https://arxiv.org/abs/2603.10062,True,core,execution_state,model,representation_organization;multi_agent_shared;multimodal_embodied_gui,personal_shared,0,agent-memory
+2604.00009,2026-03-09,transition-2026-Feb-Mar,"Eyla: Toward an Identity-Anchored LLM Architecture with Integrated Biological Priors -- Vision, Implementation Attempt, and Lessons from AI-Assisted Development",https://arxiv.org/abs/2604.00009,False,core,representation,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,,5,episodic-memory
+2603.11721,2026-03-12,transition-2026-Feb-Mar,When OpenClaw Meets Hospital: Toward an Agentic Operating System for Dynamic Clinical Workflows,https://arxiv.org/abs/2603.11721,False,core,execution_state,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,7,long-term-memory
+2603.11768,2026-03-12,transition-2026-Feb-Mar,"Governing Evolving Memory in LLM Agents: Risks, Mechanisms, and the Stability and Safety Governed Memory (SSGM) Framework",https://arxiv.org/abs/2603.11768,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;security_governance,0,agent-memory;long-term-memory
+2603.12631,2026-03-13,transition-2026-Feb-Mar,Joint Optimization of Multi-agent Memory System,https://arxiv.org/abs/2603.12631,True,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;multi_agent_shared;multimodal_embodied_gui,personal_shared,5,agent-memory
+2603.13017,2026-03-13,transition-2026-Feb-Mar,Structured Distillation for Personalized Agent Memory: 11x Token Reduction with Retrieval Preservation,https://arxiv.org/abs/2603.13017,True,core,access,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,retrieval_representation;experience_skill_policy;systems_efficiency;personal_shared,8,agent-memory
+2603.13676,2026-03-14,transition-2026-Feb-Mar,TheraAgent: Multi-Agent Framework with Self-Evolving Memory and Evidence-Calibrated Reasoning for PET Theranostics,https://arxiv.org/abs/2603.13676,True,core,execution_state,model,representation_organization;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared,experience_skill_policy;personal_shared,7,agent-memory
+2603.14588,2026-03-15,transition-2026-Feb-Mar,SuperLocalMemory V3: Information-Geometric Foundations for Zero-LLM Enterprise Agent Memory,https://arxiv.org/abs/2603.14588,True,core,access,model,retrieval_access;representation_organization;evaluation_diagnosis,,6,agent-memory
+2603.14597,2026-03-15,transition-2026-Feb-Mar,D-MEM: Dopamine-Gated Agentic Memory via Reward Prediction Error Routing,https://arxiv.org/abs/2603.14597,True,core,update_forgetting,model,representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,7,agent-memory;long-term-memory
+2603.15280,2026-03-16,transition-2026-Feb-Mar,Advancing Multimodal Agent Reasoning with Long-Term Neuro-Symbolic Memory,https://arxiv.org/abs/2603.15280,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,5,agent-memory
+2603.15421,2026-03-16,transition-2026-Feb-Mar,CLAG: Adaptive Memory Organization via Agent-Driven Clustering for Small Language Model Agents,https://arxiv.org/abs/2603.15421,True,core,access,model,retrieval_access;representation_organization;experience_skill_learning;systems_efficiency,retrieval_representation;experience_skill_policy,2,agent-memory
+2603.16734,2026-03-17,transition-2026-Feb-Mar,Differential Harm Propensity in Personalized LLM Agents: The Curious Case of Mental Health Disclosure,https://arxiv.org/abs/2603.16734,False,core,execution_state,model,evaluation_diagnosis;security_privacy_trust,personal_shared,4,long-term-memory
+2603.17043,2026-03-17,transition-2026-Feb-Mar,OpenQlaw: An Agentic AI Assistant for Analysis of 2D Quantum Materials,https://arxiv.org/abs/2603.17043,False,core,execution_state,model,retrieval_access;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis,0,long-term-memory
+2603.17244,2026-03-18,transition-2026-Feb-Mar,Graph-Native Cognitive Memory for AI Agents: Formal Belief Revision Semantics for Versioned Memory Architectures,https://arxiv.org/abs/2603.17244,True,core,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation;update_forgetting,6,agent-memory;memory-evaluation
+2603.17831,2026-03-18,transition-2026-Feb-Mar,RPMS: Enhancing LLM-Based Embodied Planning through Rule-Augmented Memory Synergy,https://arxiv.org/abs/2603.17831,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,6,episodic-memory
+2603.17948,2026-03-18,transition-2026-Feb-Mar,VideoAtlas: Navigating Long-Form Video in Logarithmic Compute,https://arxiv.org/abs/2603.17948,False,core,experience_learning,model,representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,2,agent-memory
+2603.19935,2026-03-20,transition-2026-Feb-Mar,"Memori: A Persistent Memory Layer for Efficient, Context-Aware LLM Agents",https://arxiv.org/abs/2603.19935,True,core,experience_learning,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,systems_efficiency,6,long-term-memory
+2604.14158,2026-03-23,transition-2026-Feb-Mar,MemGround: Long-Term Memory Evaluation Kit for Large Language Models in Gamified Scenarios,https://arxiv.org/abs/2604.14158,True,supporting,evaluation,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis,benchmark_diagnosis,3,agent-memory
+2605.20189,2026-03-23,transition-2026-Feb-Mar,SOLAR: A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation,https://arxiv.org/abs/2605.20189,False,supporting,evaluation,model,update_forgetting;experience_skill_learning;systems_efficiency,experience_skill_policy,6,episodic-memory
+2603.23064,2026-03-24,transition-2026-Feb-Mar,Mind Your HEARTBEAT! Claw Background Execution Inherently Enables Silent Memory Pollution,https://arxiv.org/abs/2603.23064,True,peripheral,security_privacy,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,,3,agent-memory
+2603.23160,2026-03-24,transition-2026-Feb-Mar,UniDial-EvalKit: A Unified Toolkit for Evaluating Multi-Faceted Conversational Abilities,https://arxiv.org/abs/2603.23160,False,supporting,evaluation,model,representation_organization;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis,6,agent-memory
+2603.23231,2026-03-24,transition-2026-Feb-Mar,PERMA: Benchmarking Personalized Memory Agents via Event-Driven Preference and Realistic Task Environments,https://arxiv.org/abs/2603.23231,True,core,experience_learning,model,retrieval_access;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;personal_shared,5,agent-memory
+2603.24564,2026-03-25,transition-2026-Feb-Mar,"Infrastructure for Valuable, Tradable, and Verifiable Agent Memory",https://arxiv.org/abs/2603.24564,True,supporting,evaluation,model,representation_organization;security_privacy_trust;systems_efficiency,retrieval_representation,0,agent-memory
+2603.25097,2026-03-26,transition-2026-Feb-Mar,ElephantBroker: A Knowledge-Grounded Cognitive Runtime for Trustworthy AI Agents,https://arxiv.org/abs/2603.25097,False,core,evaluation,model,retrieval_access;representation_organization;update_forgetting;security_privacy_trust;systems_efficiency,security_governance;systems_efficiency,0,agent-memory
+2603.25973,2026-03-26,transition-2026-Feb-Mar,MemoryCD: Benchmarking Long-Context User Memory of LLM Agents for Lifelong Cross-Domain Personalization,https://arxiv.org/abs/2603.25973,True,supporting,evaluation,model,evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;personal_shared,5,memory-evaluation
+2604.19771,2026-03-27,transition-2026-Feb-Mar,Cognis: Context-Aware Memory for Conversational AI Agents,https://arxiv.org/abs/2604.19771,True,core,experience_learning,model,retrieval_access;evaluation_diagnosis;systems_efficiency,,4,long-term-memory
+2603.28088,2026-03-30,transition-2026-Feb-Mar,GEMS: Agent-Native Multimodal Generation with Memory and Skills,https://arxiv.org/abs/2603.28088,True,core,experience_learning,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui,experience_skill_policy;execution_multimodal,4,agent-memory
+2603.29493,2026-03-31,transition-2026-Feb-Mar,MemFactory: Unified Inference & Training Framework for Agent Memory,https://arxiv.org/abs/2603.29493,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis,,6,agent-memory
+2604.00131,2026-03-31,transition-2026-Feb-Mar,Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven Activation,https://arxiv.org/abs/2604.00131,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,update_forgetting;experience_skill_policy,2,agent-memory
+2604.00556,2026-04-01,recent-2026-Apr-Jul,HabitatAgent: An End-to-End Multi-Agent System for Housing Consultation,https://arxiv.org/abs/2604.00556,False,core,experience_learning,model,retrieval_access;update_forgetting;evaluation_diagnosis;multi_agent_shared,personal_shared,6,agent-memory
+2604.01007,2026-04-01,recent-2026-Apr-Jul,Omni-SimpleMem: Autoresearch-Guided Discovery of Lifelong Multimodal Agent Memory,https://arxiv.org/abs/2604.01007,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,retrieval_representation;execution_multimodal,8,agent-memory
+2604.01560,2026-04-02,recent-2026-Apr-Jul,DeltaMem: Towards Agentic Memory Management via Reinforcement Learning,https://arxiv.org/abs/2604.01560,True,core,update_forgetting,model,experience_skill_learning;evaluation_diagnosis;multi_agent_shared,experience_skill_policy,4,agent-memory;memory-evaluation
+2604.01658,2026-04-02,recent-2026-Apr-Jul,CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery,https://arxiv.org/abs/2604.01658,False,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multi_agent_shared,personal_shared,5,long-term-memory
+2604.01707,2026-04-02,recent-2026-Apr-Jul,Memory in the LLM Era: Modular Architectures and Strategies in a Unified Framework,https://arxiv.org/abs/2604.01707,True,core,access,model,retrieval_access;execution_state;evaluation_diagnosis,,4,agent-memory
+2604.02522,2026-04-02,recent-2026-Apr-Jul,Opal: Private Memory for Personal AI,https://arxiv.org/abs/2604.02522,True,core,access,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,6,agent-memory
+2604.02623,2026-04-03,recent-2026-Apr-Jul,"Poison Once, Exploit Forever: Environment-Injected Memory Poisoning Attacks on Web Agents",https://arxiv.org/abs/2604.02623,True,core,update_forgetting,model,retrieval_access;execution_state;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,execution_multimodal;security_governance,4,agent-memory
+2604.03588,2026-04-04,recent-2026-Apr-Jul,Rashomon Memory: Towards Argumentation-Driven Retrieval for Multi-Perspective Agent Memory,https://arxiv.org/abs/2604.03588,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation,0,agent-memory
+2604.04157,2026-04-05,recent-2026-Apr-Jul,Readable Minds: Emergent Theory-of-Mind-Like Behavior in LLM Poker Agents,https://arxiv.org/abs/2604.04157,False,core,execution_state,model,,,8,long-term-memory
+2604.04503,2026-04-06,recent-2026-Apr-Jul,Memory Intelligence Agent,https://arxiv.org/abs/2604.04503,True,core,update_forgetting,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,3,agent-memory
+2604.04514,2026-04-06,recent-2026-Apr-Jul,"SuperLocalMemory V3.3: The Living Brain -- Biologically-Inspired Forgetting, Cognitive Quantization, and Multi-Channel Retrieval for Zero-LLM Agent Memory Systems",https://arxiv.org/abs/2604.04514,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation;update_forgetting,4,agent-memory
+2604.04660,2026-04-06,recent-2026-Apr-Jul,"Springdrift: An Auditable Persistent Runtime for LLM Agents with Case-Based Memory, Normative Safety, and Ambient Self-Perception",https://arxiv.org/abs/2604.04660,True,supporting,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui,systems_efficiency,6,long-term-memory
+2604.04853,2026-04-06,recent-2026-Apr-Jul,MemMachine: A Ground-Truth-Preserving Memory System for Personalized AI Agents,https://arxiv.org/abs/2604.04853,True,core,experience_learning,model,retrieval_access;execution_state;evaluation_diagnosis;systems_efficiency,systems_efficiency;personal_shared,6,long-term-memory
+2604.04901,2026-04-06,recent-2026-Apr-Jul,FileGram: Grounding Agent Personalization in File-System Behavioral Traces,https://arxiv.org/abs/2604.04901,False,supporting,execution_state,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,personal_shared,2,agent-memory
+2604.05719,2026-04-07,recent-2026-Apr-Jul,Hackers or Hallucinators? A Comprehensive Analysis of LLM-Based Automated Penetration Testing,https://arxiv.org/abs/2604.05719,False,supporting,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,benchmark_diagnosis,5,agent-memory
+2604.07798,2026-04-09,recent-2026-Apr-Jul,Lightweight LLM Agent Memory with Small Language Models,https://arxiv.org/abs/2604.07798,True,core,experience_learning,model,retrieval_access;update_forgetting;execution_state;systems_efficiency,,2,agent-memory;long-term-memory
+2604.07877,2026-04-09,recent-2026-Apr-Jul,MemReader: From Passive to Active Extraction for Long-Term Agent Memory,https://arxiv.org/abs/2604.07877,True,core,experience_learning,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,5,agent-memory
+2604.08064,2026-04-09,recent-2026-Apr-Jul,ImplicitMemBench: Measuring Unconscious Behavioral Adaptation in Large Language Models,https://arxiv.org/abs/2604.08064,False,supporting,execution_state,model,retrieval_access;experience_skill_learning;evaluation_diagnosis,experience_skill_policy,8,memory-evaluation;procedural-memory
+2604.08756,2026-04-09,recent-2026-Apr-Jul,Artifacts as Memory Beyond the Agent Boundary,https://arxiv.org/abs/2604.08756,True,supporting,execution_state,model,experience_skill_learning,,2,agent-memory
+2604.09000,2026-04-10,recent-2026-Apr-Jul,StreamMeCo: Long-Term Agent Memory Compression for Efficient Streaming Video Understanding,https://arxiv.org/abs/2604.09000,True,core,experience_learning,model,retrieval_access;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal;systems_efficiency,5,agent-memory
+2604.09388,2026-04-10,recent-2026-Apr-Jul,The AI Codebase Maturity Model: From Assisted Coding to Fully Autonomous Systems,https://arxiv.org/abs/2604.09388,False,core,experience_learning,model,representation_organization;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;systems_efficiency,,3,agent-memory
+2604.09747,2026-04-10,recent-2026-Apr-Jul,ADAM: A Systematic Data Extraction Attack on Agent Memory via Adaptive Querying,https://arxiv.org/abs/2604.09747,True,supporting,execution_state,model,retrieval_access;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;security_governance,6,agent-memory
+2604.11811,2026-04-10,recent-2026-Apr-Jul,M$^\star$: Every Task Deserves Its Own Memory Harness,https://arxiv.org/abs/2604.11811,True,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,5,agent-memory
+2604.10981,2026-04-13,recent-2026-Apr-Jul,"ATANT v1.1: Positioning Continuity Evaluation Against Memory, Long-Context, and Agentic-Memory Benchmarks",https://arxiv.org/abs/2604.10981,True,supporting,evaluation,model,evaluation_diagnosis;security_privacy_trust,benchmark_diagnosis,5,agent-memory;memory-evaluation
+2604.11544,2026-04-13,recent-2026-Apr-Jul,Time is Not a Label: Continuous Phase Rotation for Temporal Knowledge Graphs and Agentic Memory,https://arxiv.org/abs/2604.11544,True,core,update_forgetting,model,representation_organization;evaluation_diagnosis;multimodal_embodied_gui,retrieval_representation;systems_efficiency,4,agent-memory
+2604.11563,2026-04-13,recent-2026-Apr-Jul,Synthius-Mem: Brain-Inspired Hallucination-Resistant Persona Memory Achieving 94.4% Memory Accuracy and 99.6% Adversarial Robustness on LoCoMo,https://arxiv.org/abs/2604.11563,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,8,long-term-memory
+2604.12007,2026-04-13,recent-2026-Apr-Jul,When to Forget: A Memory Governance Primitive,https://arxiv.org/abs/2604.12007,True,core,update_forgetting,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust,update_forgetting;security_governance,6,agent-memory
+2604.12179,2026-04-14,recent-2026-Apr-Jul,AgenticAI-DialogGen: Topic-Guided Conversation Generation for Fine-Tuning and Evaluating Short- and Long-Term Memories of LLMs,https://arxiv.org/abs/2604.12179,True,core,update_forgetting,model,representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;execution_multimodal,2,long-term-memory
+2604.12285,2026-04-14,recent-2026-Apr-Jul,GAM: Hierarchical Graph-based Agentic Memory for LLM Agents,https://arxiv.org/abs/2604.12285,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;multimodal_embodied_gui;systems_efficiency,retrieval_representation,4,agent-memory
+2604.12948,2026-04-14,recent-2026-Apr-Jul,Drawing on Memory: Dual-Trace Encoding Improves Cross-Session Recall in LLM Agents,https://arxiv.org/abs/2604.12948,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;systems_efficiency,retrieval_representation,7,long-term-memory
+2604.15774,2026-04-17,recent-2026-Apr-Jul,MemEvoBench: Benchmarking Safety Risks from Memory Misevolution in LLM Agents,https://arxiv.org/abs/2604.15774,True,supporting,evaluation,model,update_forgetting;execution_state;evaluation_diagnosis;security_privacy_trust,benchmark_diagnosis,4,long-term-memory
+2604.15877,2026-04-17,recent-2026-Apr-Jul,"Experience Compression Spectrum: Unifying Memory, Skills, and Rules in LLM Agents",https://arxiv.org/abs/2604.15877,True,supporting,evaluation,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy;systems_efficiency,2,agent-memory;episodic-memory
+2604.16548,2026-04-17,recent-2026-Apr-Jul,"A Survey on Long-Term Memory Security in LLM Agents: Attacks, Defenses, and Governance Across the Memory Lifecycle",https://arxiv.org/abs/2604.16548,True,core,update_forgetting,model,retrieval_access;update_forgetting;security_privacy_trust;systems_efficiency,benchmark_diagnosis;update_forgetting;security_governance,0,long-term-memory
+2604.16839,2026-04-18,recent-2026-Apr-Jul,HeLa-Mem: Hebbian Learning and Associative Memory for LLM Agents,https://arxiv.org/abs/2604.16839,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;systems_efficiency,experience_skill_policy,5,episodic-memory;long-term-memory
+2604.17273,2026-04-19,recent-2026-Apr-Jul,The Continuity Layer: Why Intelligence Needs an Architecture for What It Carries Forward,https://arxiv.org/abs/2604.17273,False,core,execution_state,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,3,agent-memory;memory-evaluation
+2604.17456,2026-04-19,recent-2026-Apr-Jul,TrafficClaw: A Generalizable LLM Agent in the Unified Physical Environment for Urban Traffic Control,https://arxiv.org/abs/2604.17456,False,core,execution_state,model,representation_organization;execution_state;experience_skill_learning,,3,long-term-memory
+2604.17658,2026-04-19,recent-2026-Apr-Jul,Towards Self-Improving Error Diagnosis in Multi-Agent Systems,https://arxiv.org/abs/2604.17658,False,core,execution_state,model,update_forgetting;evaluation_diagnosis;multi_agent_shared,benchmark_diagnosis;personal_shared,6,episodic-memory
+2604.19541,2026-04-21,recent-2026-Apr-Jul,FOCAL: Filtered On-device Continuous Activity Logging for Efficient Personal Desktop Summarization,https://arxiv.org/abs/2604.19541,False,core,execution_state,model,retrieval_access;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,systems_efficiency,6,agent-memory
+2604.20006,2026-04-21,recent-2026-Apr-Jul,From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents,https://arxiv.org/abs/2604.20006,True,core,execution_state,model,retrieval_access;update_forgetting;evaluation_diagnosis,benchmark_diagnosis;retrieval_representation;update_forgetting;personal_shared,3,agent-memory
+2604.20117,2026-04-22,recent-2026-Apr-Jul,To Know is to Construct: Schema-Constrained Generation for Agent Memory,https://arxiv.org/abs/2604.20117,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui,,5,agent-memory
+2604.20183,2026-04-22,recent-2026-Apr-Jul,Dual-Cluster Memory Agent: Resolving Multi-Paradigm Ambiguity in Optimization Problem Solving,https://arxiv.org/abs/2604.20183,True,core,execution_state,model,representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,4,agent-memory
+2604.20300,2026-04-22,recent-2026-Apr-Jul,FSFM: A Biologically-Inspired Framework for Selective Forgetting of Agent Memory,https://arxiv.org/abs/2604.20300,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust,update_forgetting,5,agent-memory
+2604.20582,2026-04-22,recent-2026-Apr-Jul,"Trust, Lies, and Long Memories: Emergent Social Dynamics and Reputation in Multi-Round Avalon with LLM Agents",https://arxiv.org/abs/2604.20582,True,core,execution_state,model,security_privacy_trust,security_governance,4,agent-memory
+2604.20598,2026-04-22,recent-2026-Apr-Jul,"Self-Aware Vector Embeddings for Retrieval-Augmented Generation: A Neuroscience-Inspired Framework for Temporal, Confidence-Weighted, and Relational Knowledge",https://arxiv.org/abs/2604.20598,False,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation;systems_efficiency,6,agent-memory
+2604.22085,2026-04-23,recent-2026-Apr-Jul,Memanto: Typed Semantic Memory with Information-Theoretic Retrieval for Long-Horizon Agents,https://arxiv.org/abs/2604.22085,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis;systems_efficiency,retrieval_representation;execution_multimodal,3,agent-memory
+2604.23711,2026-04-26,recent-2026-Apr-Jul,Spore: Efficient and Training-Free Privacy Extraction Attack on LLMs via Inference-Time Hybrid Probing,https://arxiv.org/abs/2604.23711,False,core,execution_state,model,retrieval_access;evaluation_diagnosis;security_privacy_trust;systems_efficiency,security_governance;systems_efficiency,4,agent-memory
+2606.20570,2026-04-26,recent-2026-Apr-Jul,Infrastructure for the Agentic Web: Gap Analysis and Architecture from the Agentverse Platform,https://arxiv.org/abs/2606.20570,False,core,access,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;retrieval_representation,1,agent-memory
+2604.26197,2026-04-29,recent-2026-Apr-Jul,Hierarchical Long-Term Semantic Memory for LinkedIn's Hiring Agent,https://arxiv.org/abs/2604.26197,True,core,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation,5,long-term-memory
+2604.26622,2026-04-29,recent-2026-Apr-Jul,OCR-Memory: Optical Context Retrieval for Long-Horizon Agent Memory,https://arxiv.org/abs/2604.26622,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,retrieval_representation;execution_multimodal,3,agent-memory
+2604.27003,2026-04-29,recent-2026-Apr-Jul,When Continual Learning Moves to Memory: A Study of Experience Reuse in LLM Agents,https://arxiv.org/abs/2604.27003,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis;experience_skill_policy,3,procedural-memory
+2604.27045,2026-04-29,recent-2026-Apr-Jul,Detecting Clinical Discrepancies in Health Coaching Agents: A Dual-Stream Memory and Reconciliation Architecture,https://arxiv.org/abs/2604.27045,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust,,5,agent-memory
+2604.27283,2026-04-30,recent-2026-Apr-Jul,Learning When to Remember: Risk-Sensitive Contextual Bandits for Abstention-Aware Memory Retrieval in LLM-Based Coding Agents,https://arxiv.org/abs/2604.27283,True,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;systems_efficiency,retrieval_representation;experience_skill_policy,3,agent-memory
+2604.27707,2026-04-30,recent-2026-Apr-Jul,"Contextual Agentic Memory is a Memo, Not True Memory",https://arxiv.org/abs/2604.27707,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,3,agent-memory
+2604.27996,2026-04-30,recent-2026-Apr-Jul,Exploring LLM Agent Designs and Interaction Modalities for Scientific Visualization,https://arxiv.org/abs/2604.27996,False,core,update_forgetting,model,representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,4,long-term-memory
+2605.00702,2026-05-01,recent-2026-Apr-Jul,Learning How and What to Memorize: Cognition-Inspired Two-Stage Optimization for Evolving Memory,https://arxiv.org/abs/2605.00702,True,core,update_forgetting,model,representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy,4,memory-evaluation
+2605.01386,2026-05-02,recent-2026-Apr-Jul,MemORAI: Memory Organization and Retrieval via Adaptive Graph Intelligence for LLM Conversational Agents,https://arxiv.org/abs/2605.01386,True,core,representation,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation;experience_skill_policy,2,long-term-memory
+2605.01970,2026-05-03,recent-2026-Apr-Jul,Trojan Hippo: Weaponizing Agent Memory for Data Exfiltration,https://arxiv.org/abs/2605.01970,True,core,update_forgetting,model,evaluation_diagnosis;security_privacy_trust;systems_efficiency,,4,agent-memory;long-term-memory
+2605.02199,2026-05-04,recent-2026-Apr-Jul,MEMAUDIT: An Exact Package-Oracle Evaluation Protocol for Budgeted Long-Term LLM Memory Writing,https://arxiv.org/abs/2605.02199,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;systems_efficiency,3,long-term-memory
+2605.03228,2026-05-04,recent-2026-Apr-Jul,MAGE: Safeguarding LLM Agents against Long-Horizon Threats via Shadow Memory,https://arxiv.org/abs/2605.03228,True,core,update_forgetting,model,retrieval_access;execution_state;evaluation_diagnosis;security_privacy_trust,execution_multimodal,5,agent-memory
+2605.03312,2026-05-05,recent-2026-Apr-Jul,MemFlow: Intent-Driven Memory Orchestration for Small Language Model Agents,https://arxiv.org/abs/2605.03312,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;systems_efficiency,,7,agent-memory;memory-evaluation
+2605.03354,2026-05-05,recent-2026-Apr-Jul,What Happens Inside Agent Memory? Circuit Analysis from Emergence to Diagnosis,https://arxiv.org/abs/2605.03354,True,core,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis,5,agent-memory
+2605.03804,2026-05-05,recent-2026-Apr-Jul,ScrapMem: A Bio-inspired Framework for On-device Personalized Agent Memory via Optical Forgetting,https://arxiv.org/abs/2605.03804,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;multimodal_embodied_gui;systems_efficiency,update_forgetting;systems_efficiency;personal_shared,4,agent-memory;episodic-memory;long-term-memory
+2605.06702,2026-05-05,recent-2026-Apr-Jul,CASCADE: Case-Based Continual Adaptation for Large Language Models During Deployment,https://arxiv.org/abs/2605.06702,False,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy,5,episodic-memory
+2605.04811,2026-05-06,recent-2026-Apr-Jul,Tree-based Credit Assignment for Multi-Agent Memory System,https://arxiv.org/abs/2605.04811,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;systems_efficiency,experience_skill_policy;personal_shared,5,agent-memory
+2605.04897,2026-05-06,recent-2026-Apr-Jul,Storage Is Not Memory: A Retrieval-Centered Architecture for Agent Recall,https://arxiv.org/abs/2605.04897,True,core,access,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,retrieval_representation;systems_efficiency,3,agent-memory
+2605.05583,2026-05-07,recent-2026-Apr-Jul,Belief Memory: Agent Memory Under Partial Observability,https://arxiv.org/abs/2605.05583,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis,,5,agent-memory
+2605.06132,2026-05-07,recent-2026-Apr-Jul,MemReranker: Reasoning-Aware Reranking for Agent Memory Retrieval,https://arxiv.org/abs/2605.06132,True,core,update_forgetting,model,retrieval_access;evaluation_diagnosis;systems_efficiency,retrieval_representation,7,agent-memory
+2605.06527,2026-05-07,recent-2026-Apr-Jul,STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?,https://arxiv.org/abs/2605.06527,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,update_forgetting,7,agent-memory
+2605.06716,2026-05-07,recent-2026-Apr-Jul,From Storage to Experience: A Survey on the Evolution of LLM Agent Memory Mechanisms,https://arxiv.org/abs/2605.06716,True,core,experience_learning,model,retrieval_access;execution_state;experience_skill_learning;systems_efficiency,benchmark_diagnosis;experience_skill_policy;systems_efficiency,0,agent-memory
+2605.06812,2026-05-07,recent-2026-Apr-Jul,Towards Security-Auditable LLM Agents: A Unified Graph Representation,https://arxiv.org/abs/2605.06812,False,core,security_privacy,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;multi_agent_shared,retrieval_representation;security_governance,3,long-term-memory
+2605.07242,2026-05-08,recent-2026-Apr-Jul,MEMOREPAIR: Barrier-First Cascade Repair in Agentic Memory,https://arxiv.org/abs/2605.07242,True,core,update_forgetting,model,update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,8,agent-memory
+2605.07313,2026-05-08,recent-2026-Apr-Jul,When Stored Evidence Stops Being Usable: Scale-Conditioned Evaluation of Agent Memory,https://arxiv.org/abs/2605.07313,True,supporting,evaluation,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis,3,agent-memory
+2605.08374,2026-05-08,recent-2026-Apr-Jul,MemQ: Integrating Q-Learning into Self-Evolving Memory Agents over Provenance DAGs,https://arxiv.org/abs/2605.08374,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,experience_skill_policy;security_governance,3,agent-memory;episodic-memory
+2605.08442,2026-05-08,recent-2026-Apr-Jul,Defense effectiveness across architectural layers: a mechanistic evaluation of persistent memory attacks on stateful LLM agents,https://arxiv.org/abs/2605.08442,True,supporting,security_privacy,model,retrieval_access;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;security_governance,7,long-term-memory
+2605.08468,2026-05-08,recent-2026-Apr-Jul,"PYTHALAB-MERA: Validation-Grounded Memory, Retrieval, and Acceptance Control for Frozen-LLM Coding Agents",https://arxiv.org/abs/2605.08468,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;systems_efficiency,retrieval_representation,5,episodic-memory
+2605.08538,2026-05-08,recent-2026-Apr-Jul,Human-Inspired Memory Architecture for LLM Agents,https://arxiv.org/abs/2605.08538,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,8,long-term-memory
+2605.09033,2026-05-09,recent-2026-Apr-Jul,ShadowMerge: A Novel Poisoning Attack on Graph-Based Agent Memory via Relation-Channel Conflicts,https://arxiv.org/abs/2605.09033,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust,retrieval_representation;security_governance,7,agent-memory
+2605.09315,2026-05-10,recent-2026-Apr-Jul,Do Self-Evolving Agents Forget? Capability Degradation and Preservation in Lifelong LLM Agent Adaptation,https://arxiv.org/abs/2605.09315,True,core,update_forgetting,model,update_forgetting;execution_state;experience_skill_learning;systems_efficiency,update_forgetting;experience_skill_policy,0,long-term-memory
+2605.09330,2026-05-10,recent-2026-Apr-Jul,The Trap of Trajectory: Towards Understanding and Mitigating Spurious Correlations in Agentic Memory,https://arxiv.org/abs/2605.09330,True,supporting,evaluation,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,2,agent-memory
+2605.11032,2026-05-10,recent-2026-Apr-Jul,Portable Agent Memory: A Protocol for Cryptographically-Verified Memory Transfer Across Heterogeneous AI Agents,https://arxiv.org/abs/2605.11032,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;security_privacy_trust;systems_efficiency,retrieval_representation,1,agent-memory
+2605.09863,2026-05-11,recent-2026-Apr-Jul,Nautilus Compass: Black-box Persona Drift Detection for Production LLM Agents,https://arxiv.org/abs/2605.09863,False,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;systems_efficiency,,7,agent-memory
+2605.09942,2026-05-11,recent-2026-Apr-Jul,HAGE: Harnessing Agentic Memory via RL-Driven Weighted Graph Evolution,https://arxiv.org/abs/2605.09942,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;multi_agent_shared,retrieval_representation;experience_skill_policy,3,agent-memory
+2605.10268,2026-05-11,recent-2026-Apr-Jul,MemReread: Enhancing Agentic Long-Context Reasoning via Memory-Guided Rereading,https://arxiv.org/abs/2605.10268,True,core,update_forgetting,model,retrieval_access;update_forgetting;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,execution_multimodal,5,agent-memory
+2605.10870,2026-05-11,recent-2026-Apr-Jul,"Remember the Decision, Not the Description: A Rate-Distortion Framework for Agent Memory",https://arxiv.org/abs/2605.10870,True,core,update_forgetting,model,update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,,3,agent-memory
+2605.10899,2026-05-11,recent-2026-Apr-Jul,RubricEM: Meta-RL with Rubric-guided Policy Decomposition beyond Verifiable Rewards,https://arxiv.org/abs/2605.10899,False,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy;execution_multimodal,4,agent-memory
+2605.13880,2026-05-11,recent-2026-Apr-Jul,PREPING: Building Agent Memory without Tasks,https://arxiv.org/abs/2605.13880,True,core,update_forgetting,model,representation_organization;update_forgetting;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,,5,agent-memory
+2605.11814,2026-05-12,recent-2026-Apr-Jul,MedMemoryBench: Benchmarking Agent Memory in Personalized Healthcare,https://arxiv.org/abs/2605.11814,True,core,execution_state,model,retrieval_access;execution_state;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;personal_shared,3,agent-memory
+2605.12061,2026-05-12,recent-2026-Apr-Jul,SAGE: A Self-Evolving Agentic Graph-Memory Engine for Structure-Aware Associative Memory,https://arxiv.org/abs/2605.12061,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis,retrieval_representation;experience_skill_policy,2,agent-memory;long-term-memory;memory-evaluation
+2605.12213,2026-05-12,recent-2026-Apr-Jul,Goal-Oriented Reasoning for RAG-based Memory in Conversational Agentic LLM Systems,https://arxiv.org/abs/2605.12213,True,core,update_forgetting,model,retrieval_access;execution_state,retrieval_representation,5,agent-memory
+2605.12294,2026-05-12,recent-2026-Apr-Jul,Executable Agentic Memory for GUI Agent,https://arxiv.org/abs/2605.12294,True,core,execution_state,model,retrieval_access;representation_organization;execution_state;multimodal_embodied_gui;systems_efficiency,execution_multimodal,2,agent-memory
+2605.12493,2026-05-12,recent-2026-Apr-Jul,LongMemEval-V2: Evaluating Long-Term Agent Memory Toward Experienced Colleagues,https://arxiv.org/abs/2605.12493,True,core,experience_learning,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;experience_skill_policy,8,agent-memory;memory-evaluation
+2605.12978,2026-05-13,recent-2026-Apr-Jul,Useful Memories Become Faulty When Continuously Updated by LLMs,https://arxiv.org/abs/2605.12978,True,core,update_forgetting,model,update_forgetting;experience_skill_learning;evaluation_diagnosis,update_forgetting,6,agent-memory
+2605.13438,2026-05-13,recent-2026-Apr-Jul,CogniFold: Always-On Proactive Memory via Cognitive Folding,https://arxiv.org/abs/2605.13438,True,core,representation,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis,,3,agent-memory
+2605.13542,2026-05-13,recent-2026-Apr-Jul,RealICU: Do LLM Agents Understand Long-Context ICU Data? A Benchmark Beyond Behavior Imitation,https://arxiv.org/abs/2605.13542,False,core,execution_state,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis,benchmark_diagnosis,4,agent-memory
+2605.13941,2026-05-13,recent-2026-Apr-Jul,EvolveMem:Self-Evolving Memory Architecture via AutoResearch for LLM Agents,https://arxiv.org/abs/2605.13941,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis,retrieval_representation;experience_skill_policy,8,long-term-memory
+2605.14421,2026-05-14,recent-2026-Apr-Jul,MemLineage: Lineage-Guided Enforcement for LLM Agent Memory,https://arxiv.org/abs/2605.14421,True,supporting,security_privacy,model,retrieval_access;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,execution_multimodal,5,agent-memory
+2605.14498,2026-05-14,recent-2026-Apr-Jul,GroupMemBench: Benchmarking LLM Agent Memory in Multi-Party Conversations,https://arxiv.org/abs/2605.14498,True,supporting,evaluation,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,benchmark_diagnosis,7,agent-memory
+2605.14906,2026-05-14,recent-2026-Apr-Jul,MemLens: Benchmarking Multimodal Long-Term Memory in Large Vision-Language Models,https://arxiv.org/abs/2605.14906,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;execution_multimodal,6,agent-memory
+2605.15128,2026-05-14,recent-2026-Apr-Jul,MemEye: A Visual-Centric Evaluation Framework for Multimodal Agent Memory,https://arxiv.org/abs/2605.15128,True,supporting,evaluation,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;execution_multimodal,3,agent-memory
+2605.15338,2026-05-14,recent-2026-Apr-Jul,Hidden in Memory: Sleeper Memory Poisoning in LLM Agents,https://arxiv.org/abs/2605.15338,True,supporting,security_privacy,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,security_governance,4,long-term-memory
+2605.15701,2026-05-15,recent-2026-Apr-Jul,H-Mem: A Novel Memory Mechanism for Evolving and Retrieving Agent Memory via a Hybrid Structure,https://arxiv.org/abs/2605.15701,True,core,update_forgetting,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,retrieval_representation;experience_skill_policy,2,agent-memory;memory-evaluation
+2605.15710,2026-05-15,recent-2026-Apr-Jul,SMMBench: A Benchmark for Source-Distributed Multimodal Agent Memory,https://arxiv.org/abs/2605.15710,True,supporting,evaluation,model,retrieval_access;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;execution_multimodal,4,agent-memory;memory-evaluation
+2605.15759,2026-05-15,recent-2026-Apr-Jul,DimMem: Dimensional Structuring for Efficient Long-Term Agent Memory,https://arxiv.org/abs/2605.15759,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,systems_efficiency,5,agent-memory;long-term-memory
+2605.16233,2026-05-15,recent-2026-Apr-Jul,FORGE: Self-Evolving Agent Memory With No Weight Updates via Population Broadcast,https://arxiv.org/abs/2605.16233,True,supporting,evaluation,model,representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,update_forgetting;experience_skill_policy,8,agent-memory
+2605.16481,2026-05-15,recent-2026-Apr-Jul,"Visual Agentic Memory: Enabling Online Long Video Understanding via Online Indexing, Hierarchical Memory, and Agentic Retrieval",https://arxiv.org/abs/2605.16481,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;multimodal_embodied_gui,retrieval_representation;execution_multimodal,5,agent-memory
+2605.16746,2026-05-16,recent-2026-Apr-Jul,State Contamination in Memory-Augmented LLM Agents,https://arxiv.org/abs/2605.16746,True,supporting,security_privacy,model,retrieval_access;representation_organization;update_forgetting;execution_state;security_privacy_trust;multi_agent_shared;systems_efficiency,security_governance,6,long-term-memory
+2605.16858,2026-05-16,recent-2026-Apr-Jul,Pedestrian-Aware LLM-Driven Behavioral Planning for Autonomous Vehicles,https://arxiv.org/abs/2605.16858,False,core,execution_state,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis,,7,episodic-memory
+2605.23986,2026-05-16,recent-2026-Apr-Jul,MemForest: An Efficient Agent Memory System with Hierarchical Temporal Indexing,https://arxiv.org/abs/2605.23986,True,core,update_forgetting,model,representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,retrieval_representation;systems_efficiency,8,agent-memory;memory-evaluation
+2605.17348,2026-05-17,recent-2026-Apr-Jul,"Taming ""Zombie'' Agents: A Markov State-Aware Framework for Resilient Multi-Agent Evolution",https://arxiv.org/abs/2605.17348,False,core,update_forgetting,model,multi_agent_shared;systems_efficiency,personal_shared,5,agent-memory
+2605.17596,2026-05-17,recent-2026-Apr-Jul,"NeuSymMS: A Hybrid Neuro-Symbolic Memory System for Persistent, Self-Curating LLM Agents",https://arxiv.org/abs/2605.17596,True,core,update_forgetting,model,retrieval_access;representation_organization;security_privacy_trust,,0,long-term-memory
+2605.17625,2026-05-17,recent-2026-Apr-Jul,Episodic-Semantic Memory Architecture for Long-Horizon Scientific Agents,https://arxiv.org/abs/2605.17625,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;evaluation_diagnosis;systems_efficiency,execution_multimodal,7,agent-memory
+2605.17641,2026-05-17,recent-2026-Apr-Jul,Causal Intervention-Based Memory Selection for Long-Horizon LLM Agents,https://arxiv.org/abs/2605.17641,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis,execution_multimodal,6,long-term-memory
+2605.18284,2026-05-18,recent-2026-Apr-Jul,CommitDistill: A Lightweight Knowledge-Centric Memory Layer for Software Repositories,https://arxiv.org/abs/2605.18284,True,core,execution_state,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;systems_efficiency,6,agent-memory
+2605.18421,2026-05-18,recent-2026-Apr-Jul,EvoMemBench: Benchmarking Agent Memory from a Self-Evolving Perspective,https://arxiv.org/abs/2605.18421,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis;experience_skill_policy,6,agent-memory;long-term-memory
+2605.18652,2026-05-18,recent-2026-Apr-Jul,MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents,https://arxiv.org/abs/2605.18652,True,core,update_forgetting,model,retrieval_access;execution_state;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;execution_multimodal,5,agent-memory
+2605.18930,2026-05-18,recent-2026-Apr-Jul,OEP: Poisoning Self-Evolving LLM Agents via Locally Correct but Non-Transferable Experiences,https://arxiv.org/abs/2605.18930,False,core,update_forgetting,model,representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust,experience_skill_policy;security_governance,6,agent-memory
+2605.19952,2026-05-19,recent-2026-Apr-Jul,Rethinking How to Remember: Beyond Atomic Facts in Lifelong LLM Agent Memory,https://arxiv.org/abs/2605.19952,True,core,update_forgetting,model,retrieval_access;representation_organization;systems_efficiency,benchmark_diagnosis,6,agent-memory
+2605.20616,2026-05-20,recent-2026-Apr-Jul,Auto-Dreamer: Learning Offline Memory Consolidation for Language Agents,https://arxiv.org/abs/2605.20616,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;security_privacy_trust,update_forgetting;experience_skill_policy,4,agent-memory
+2605.20724,2026-05-20,recent-2026-Apr-Jul,CALMem : Application-Layer Dual Memory for Conversational AI,https://arxiv.org/abs/2605.20724,True,core,execution_state,model,retrieval_access;representation_organization;multimodal_embodied_gui;systems_efficiency,,0,episodic-memory
+2605.20833,2026-05-20,recent-2026-Apr-Jul,MemGym: a Long-Horizon Memory Environment for LLM Agents,https://arxiv.org/abs/2605.20833,True,core,execution_state,model,retrieval_access;update_forgetting;execution_state;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,3,agent-memory;memory-evaluation
+2605.21463,2026-05-20,recent-2026-Apr-Jul,Mem-$π$: Adaptive Memory through Learning When and What to Generate,https://arxiv.org/abs/2605.21463,True,core,execution_state,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy,6,episodic-memory
+2605.22411,2026-05-21,recent-2026-Apr-Jul,DeferMem: Query-Time Evidence Distillation via Reinforcement Learning for Long-Term Memory QA,https://arxiv.org/abs/2605.22411,True,core,execution_state,model,retrieval_access;representation_organization;experience_skill_learning;systems_efficiency,experience_skill_policy,2,long-term-memory
+2605.22721,2026-05-21,recent-2026-Apr-Jul,Self-Evolving Multi-Agent Systems via Decentralized Memory,https://arxiv.org/abs/2605.22721,True,core,execution_state,model,update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;personal_shared,6,long-term-memory
+2605.23067,2026-05-21,recent-2026-Apr-Jul,What Training Data Teaches RL Memory Agents: An Empirical Study of Curriculum Effects in Memory-Augmented QA,https://arxiv.org/abs/2605.23067,True,core,execution_state,model,experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis;experience_skill_policy,3,agent-memory
+2605.23723,2026-05-22,recent-2026-Apr-Jul,MemAudit: Post-hoc Auditing of Poisoned Agent Memory via Causal Attribution and Structural Anomaly Detection,https://arxiv.org/abs/2605.23723,True,core,execution_state,model,retrieval_access;execution_state;evaluation_diagnosis;security_privacy_trust,security_governance,2,agent-memory;long-term-memory
+2605.24941,2026-05-24,recent-2026-Apr-Jul,Memory-Induced Tool-Drift in LLM Agents,https://arxiv.org/abs/2605.24941,True,core,execution_state,model,evaluation_diagnosis;security_privacy_trust;systems_efficiency,,5,long-term-memory
+2605.25002,2026-05-24,recent-2026-Apr-Jul,MemMark: State-Evolution Attribution Watermarking for Agent Long-Term Memory Systems,https://arxiv.org/abs/2605.25002,True,core,execution_state,model,update_forgetting;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,,5,agent-memory
+2605.26154,2026-05-24,recent-2026-Apr-Jul,MemMorph: Tool Hijacking in LLM Agents via Memory Poisoning,https://arxiv.org/abs/2605.26154,True,core,execution_state,model,experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,security_governance,6,long-term-memory
+2605.25869,2026-05-25,recent-2026-Apr-Jul,Mitigating Provenance-Role Collapse in Long-Term Agents via Typed Memory Representation,https://arxiv.org/abs/2605.25869,True,core,execution_state,model,retrieval_access;representation_organization;security_privacy_trust;systems_efficiency,retrieval_representation;security_governance,5,long-term-memory
+2605.26252,2026-05-25,recent-2026-Apr-Jul,Is Agent Memory a Database? Rethinking Data Foundations for Long-Term AI Agent Memory,https://arxiv.org/abs/2605.26252,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;execution_state;security_privacy_trust;systems_efficiency,benchmark_diagnosis;retrieval_representation,1,agent-memory
+2605.26256,2026-05-25,recent-2026-Apr-Jul,Personalizing Embodied Multimodal Large Language Model Agents over Long-term User Interactions,https://arxiv.org/abs/2605.26256,False,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal;personal_shared,2,episodic-memory
+2605.27366,2026-05-26,recent-2026-Apr-Jul,"MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation",https://arxiv.org/abs/2605.27366,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis;experience_skill_policy,6,agent-memory
+2605.27760,2026-05-26,recent-2026-Apr-Jul,SkillGrad: Optimizing Agent Skills Like Gradient Descent,https://arxiv.org/abs/2605.27760,False,core,update_forgetting,model,representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis,experience_skill_policy,5,long-term-memory
+2605.27762,2026-05-26,recent-2026-Apr-Jul,PEAM: Parametric Embodied Agent Memory through Contrastive Internalization of Experience in Minecraft,https://arxiv.org/abs/2605.27762,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,experience_skill_policy;execution_multimodal,5,agent-memory
+2605.27825,2026-05-27,recent-2026-Apr-Jul,MRMMIA: Membership Inference Attacks on Memory in Chat Agents,https://arxiv.org/abs/2605.27825,True,core,security_privacy,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust,security_governance,4,agent-memory
+2605.28046,2026-05-27,recent-2026-Apr-Jul,MemCog: From Memory-as-Tool to Memory-as-Cognition in Conversational Agents,https://arxiv.org/abs/2605.28046,True,core,execution_state,model,retrieval_access;evaluation_diagnosis;multimodal_embodied_gui,,4,agent-memory
+2605.29341,2026-05-28,recent-2026-Apr-Jul,WorldMemArena: Evaluating Multimodal Agent Memory Through Action-World Interaction,https://arxiv.org/abs/2605.29341,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;execution_multimodal,5,agent-memory
+2605.29630,2026-05-28,recent-2026-Apr-Jul,Entity-Collision: A Stratified Protocol for Attributing Retrieval Lift in Agent Memory,https://arxiv.org/abs/2605.29630,True,core,update_forgetting,model,retrieval_access;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation,7,agent-memory;memory-evaluation
+2605.29640,2026-05-28,recent-2026-Apr-Jul,VikingMem: A Memory Base Management System for Stateful LLM-based Applications,https://arxiv.org/abs/2605.29640,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;systems_efficiency,,6,agent-memory;memory-evaluation
+2605.29960,2026-05-28,recent-2026-Apr-Jul,Hijacking Agent Memory: Stealthy Trojan Attacks Through Conversational Interaction,https://arxiv.org/abs/2605.29960,True,core,update_forgetting,model,evaluation_diagnosis;security_privacy_trust,security_governance,5,agent-memory;long-term-memory
+2605.30058,2026-05-28,recent-2026-Apr-Jul,HEART-Bench: Do LLM Agents Exhibit Human-like Psychology?,https://arxiv.org/abs/2605.30058,False,core,experience_learning,model,representation_organization;update_forgetting;execution_state;evaluation_diagnosis;multimodal_embodied_gui,,2,episodic-memory
+2605.30621,2026-05-28,recent-2026-Apr-Jul,Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents,https://arxiv.org/abs/2605.30621,False,core,update_forgetting,model,update_forgetting;execution_state;experience_skill_learning;systems_efficiency,experience_skill_policy,2,procedural-memory
+2607.00017,2026-05-28,recent-2026-Apr-Jul,Learning User-Aware Recall: Personalized Retrieval in Long-Term Conversational Memory,https://arxiv.org/abs/2607.00017,True,core,experience_learning,model,retrieval_access;evaluation_diagnosis;multimodal_embodied_gui,retrieval_representation;experience_skill_policy;personal_shared,5,long-term-memory
+2605.30690,2026-05-29,recent-2026-Apr-Jul,ElasticMem: Latent Memory as a Learnable Resource for LLM Agents,https://arxiv.org/abs/2605.30690,True,core,representation,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,8,long-term-memory
+2605.30711,2026-05-29,recent-2026-Apr-Jul,SAGE: A Novelty Gate for Efficient Memory Evolution in Agentic LLMs,https://arxiv.org/abs/2605.30711,True,core,representation,model,retrieval_access;systems_efficiency,systems_efficiency,3,agent-memory
+2605.30858,2026-05-29,recent-2026-Apr-Jul,ForecastCompass: Guiding Agentic Forecasting with Adaptive Factor Memory,https://arxiv.org/abs/2605.30858,True,core,representation,model,retrieval_access;representation_organization;experience_skill_learning;multimodal_embodied_gui,experience_skill_policy;execution_multimodal,2,agent-memory
+2605.31075,2026-05-29,recent-2026-Apr-Jul,Task-Focused Memorization for Multimodal Agents,https://arxiv.org/abs/2605.31075,True,core,representation,model,experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,4,agent-memory
+2606.00619,2026-05-30,recent-2026-Apr-Jul,MemPro: Agentic Memory Systems as Evolvable Programs,https://arxiv.org/abs/2606.00619,True,core,representation,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,6,agent-memory
+2606.00734,2026-05-30,recent-2026-Apr-Jul,EMA: Approximate Nearest Neighbor Search with General Attribute Filtering and Dynamic Updates,https://arxiv.org/abs/2606.00734,False,core,representation,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,retrieval_representation;update_forgetting,4,agent-memory
+2606.00756,2026-05-30,recent-2026-Apr-Jul,CoMIC: Collaborative Memory and Insights Circulation for Long-Horizon LLM Agents in Cloud-Edge Systems,https://arxiv.org/abs/2606.00756,True,core,representation,model,representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,execution_multimodal;systems_efficiency;personal_shared,2,long-term-memory
+2606.01138,2026-05-31,recent-2026-Apr-Jul,memorywire: A Vendor-Neutral Wire Format for Agent Memory Operations,https://arxiv.org/abs/2606.01138,True,core,representation,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;systems_efficiency,systems_efficiency,6,agent-memory
+2606.01528,2026-06-01,recent-2026-Apr-Jul,Joint Agent Memory and Exploration Learning via Novelty Signals,https://arxiv.org/abs/2606.01528,True,core,representation,model,evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,5,agent-memory
+2606.01722,2026-06-01,recent-2026-Apr-Jul,Post-Deterministic Distributed Systems: A New Foundation for Trustworthy Autonomous Infrastructure,https://arxiv.org/abs/2606.01722,False,core,representation,model,retrieval_access;representation_organization;security_privacy_trust;multimodal_embodied_gui,retrieval_representation;security_governance,1,agent-memory
+2606.02812,2026-06-01,recent-2026-Apr-Jul,Traj-Evolve: A Self-Evolving Multi-Agent System for Patient Trajectory Modeling in Lung Cancer Early Detection,https://arxiv.org/abs/2606.02812,False,core,representation,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui,experience_skill_policy;personal_shared,3,agent-memory
+2606.03197,2026-06-02,recent-2026-Apr-Jul,MemTrain: Self-Supervised Context Memory Training,https://arxiv.org/abs/2606.03197,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,,4,agent-memory
+2606.03329,2026-06-02,recent-2026-Apr-Jul,InfoMem: Training Long-Context Memory Agents with Answer-Conditioned Information Gain,https://arxiv.org/abs/2606.03329,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;systems_efficiency,,5,agent-memory
+2606.03374,2026-06-02,recent-2026-Apr-Jul,eMEM: A Hybrid Spatio-Temporal Memory System For Embodied Agents,https://arxiv.org/abs/2606.03374,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,6,agent-memory;memory-evaluation
+2606.04120,2026-06-02,recent-2026-Apr-Jul,SaliMory: Orchestrating Cognitive Memory for Conversational Agents,https://arxiv.org/abs/2606.04120,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning,,5,agent-memory
+2606.04315,2026-06-03,recent-2026-Apr-Jul,Exploring Cross-Scenario Generality of Agentic Memory Systems: Diagnostics and a Strong Baseline,https://arxiv.org/abs/2606.04315,True,core,execution_state,model,retrieval_access;execution_state;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis,2,agent-memory
+2606.04329,2026-06-03,recent-2026-Apr-Jul,From Untrusted Input to Trusted Memory: A Systematic Study of Memory Poisoning Attacks in LLM Agents,https://arxiv.org/abs/2606.04329,True,core,security_privacy,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,benchmark_diagnosis;security_governance,3,long-term-memory
+2606.04555,2026-06-03,recent-2026-Apr-Jul,Temporal Order Matters for Agentic Memory: Segment Trees for Long-Horizon Agents,https://arxiv.org/abs/2606.04555,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis;systems_efficiency,execution_multimodal,4,agent-memory;memory-evaluation
+2606.04628,2026-06-03,recent-2026-Apr-Jul,RAMPART: Registry-based Agentic Memory with Priority-Aware Runtime Transformation,https://arxiv.org/abs/2606.04628,True,core,update_forgetting,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;systems_efficiency,systems_efficiency,6,agent-memory
+2606.04780,2026-06-03,recent-2026-Apr-Jul,PersonaTree: Structured Lifecycle Memory for Person Understanding in LLM Agents,https://arxiv.org/abs/2606.04780,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,retrieval_representation;update_forgetting,3,agent-memory;long-term-memory;memory-evaluation
+2606.05513,2026-06-03,recent-2026-Apr-Jul,EpiEvolve: Self-Evolving Agents for Streaming Pandemic Forecasting under Regime Shifts,https://arxiv.org/abs/2606.05513,False,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;security_privacy_trust,experience_skill_policy,5,episodic-memory
+2606.28349,2026-06-03,recent-2026-Apr-Jul,HMARS: A Hierarchical Multi-Agent Memory System for Long-Context Reasoning,https://arxiv.org/abs/2606.28349,True,core,representation,model,retrieval_access;representation_organization;evaluation_diagnosis;multi_agent_shared,retrieval_representation;personal_shared,4,agent-memory
+2606.05646,2026-06-04,recent-2026-Apr-Jul,Enhancing Software Engineering Through Closed-Loop Memory Optimization,https://arxiv.org/abs/2606.05646,True,core,update_forgetting,model,experience_skill_learning;evaluation_diagnosis;systems_efficiency,,4,episodic-memory
+2606.05684,2026-06-04,recent-2026-Apr-Jul,AdaMEM: Test-Time Adaptive Memory for Language Agents,https://arxiv.org/abs/2606.05684,True,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,6,agent-memory
+2606.06054,2026-06-04,recent-2026-Apr-Jul,Beyond Similarity: Trustworthy Memory Search for Personal AI Agents,https://arxiv.org/abs/2606.06054,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation;security_governance,2,agent-memory
+2606.06090,2026-06-04,recent-2026-Apr-Jul,Beyond Semantic Organization: Memory as Execution State Management for Long-Horizon Agents,https://arxiv.org/abs/2606.06090,True,core,execution_state,model,retrieval_access;representation_organization;execution_state;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,retrieval_representation;execution_multimodal,6,agent-memory
+2606.06240,2026-06-04,recent-2026-Apr-Jul,TOKI: A Bitemporal Operator Algebra for Contradiction Resolution in LLM-Agent Persistent Memory,https://arxiv.org/abs/2606.06240,True,core,update_forgetting,model,update_forgetting;security_privacy_trust,,2,long-term-memory
+2606.06448,2026-06-04,recent-2026-Apr-Jul,Agent Memory: Characterization and System Implications of Stateful Long-Horizon Workloads,https://arxiv.org/abs/2606.06448,True,core,execution_state,model,retrieval_access;update_forgetting;execution_state;evaluation_diagnosis;systems_efficiency,execution_multimodal,2,agent-memory
+2606.07402,2026-06-05,recent-2026-Apr-Jul,M$^3$Exam: Benchmarking Multimodal Memory for Realistic User-Agent Interactions,https://arxiv.org/abs/2606.07402,True,core,access,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;execution_multimodal,4,memory-evaluation
+2606.09900,2026-06-05,recent-2026-Apr-Jul,"Less Context, More Accuracy: A Bi-Temporal Memory Engine for LLM Agents Where a Lean Retrieved Context Beats the Full History",https://arxiv.org/abs/2606.09900,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,retrieval_representation,7,long-term-memory;memory-evaluation
+2606.24896,2026-06-05,recent-2026-Apr-Jul,Why Memory Components Fail: Eight Years of License and Sustainability Events in Open-Source Data Infrastructure,https://arxiv.org/abs/2606.24896,True,core,access,model,representation_organization;experience_skill_learning;security_privacy_trust,retrieval_representation,1,agent-memory
+2606.08367,2026-06-06,recent-2026-Apr-Jul,Emergence World: A Platform for Evaluating Long-Horizon Multi-Agent Autonomy,https://arxiv.org/abs/2606.08367,False,core,execution_state,model,retrieval_access;execution_state;evaluation_diagnosis;security_privacy_trust;multi_agent_shared,benchmark_diagnosis;execution_multimodal;personal_shared,3,long-term-memory
+2606.09198,2026-06-08,recent-2026-Apr-Jul,MASS: Deep Research for Social Sciences with Memory-Augmented Social Simulation,https://arxiv.org/abs/2606.09198,True,core,experience_learning,model,retrieval_access;representation_organization;update_forgetting;execution_state;multimodal_embodied_gui,retrieval_representation,4,agent-memory
+2606.09461,2026-06-08,recent-2026-Apr-Jul,H2HMem: A Multimodal Memory Benchmark for Agents in Human-Human Interactions,https://arxiv.org/abs/2606.09461,True,core,representation,model,retrieval_access;update_forgetting;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;execution_multimodal,3,memory-evaluation
+2606.09483,2026-06-08,recent-2026-Apr-Jul,Memory Beyond Recall: A Dual-Process Cognitive Memory System for Self-Evolving LLM Agents,https://arxiv.org/abs/2606.09483,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;experience_skill_learning;evaluation_diagnosis,retrieval_representation;experience_skill_policy,2,agent-memory;long-term-memory
+2606.09774,2026-06-08,recent-2026-Apr-Jul,Auto-Configuring Scientific Simulators with Lightweight Coding-Agent Adapters,https://arxiv.org/abs/2606.09774,False,core,execution_state,model,retrieval_access;experience_skill_learning;systems_efficiency,experience_skill_policy,2,agent-memory
+2606.10062,2026-06-08,recent-2026-Apr-Jul,Deployment-Time Memorization in Foundation-Model Agents,https://arxiv.org/abs/2606.10062,True,supporting,evaluation,model,retrieval_access;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,5,agent-memory
+2606.10299,2026-06-09,recent-2026-Apr-Jul,What Spatial Memory Must Store: Occlusion as the Test for Language-Agent Memory,https://arxiv.org/abs/2606.10299,True,core,execution_state,model,retrieval_access,execution_multimodal,7,agent-memory
+2606.10423,2026-06-09,recent-2026-Apr-Jul,WebChallenger: A Reliable and Efficient Generalist Web Agent,https://arxiv.org/abs/2606.10423,False,core,execution_state,model,representation_organization;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,execution_multimodal;systems_efficiency,4,long-term-memory
+2606.10532,2026-06-09,recent-2026-Apr-Jul,ActiveMem: Distributed Active Memory for Long-Horizon LLM Reasoning,https://arxiv.org/abs/2606.10532,True,core,execution_state,model,retrieval_access;update_forgetting;execution_state,execution_multimodal,3,agent-memory
+2606.10577,2026-06-09,recent-2026-Apr-Jul,AgenticNav: Zero-Shot Vision-and-Language Navigation as a Tool-Calling Harness,https://arxiv.org/abs/2606.10577,False,core,execution_state,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,6,agent-memory
+2606.10677,2026-06-09,recent-2026-Apr-Jul,Infini Memory: Maintainable Topic Documents for Long-Term LLM Agent Memory,https://arxiv.org/abs/2606.10677,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,retrieval_representation,3,agent-memory;long-term-memory
+2606.11680,2026-06-10,recent-2026-Apr-Jul,Organize then Retrieve: Hierarchical Memory Navigation for Efficient Agents,https://arxiv.org/abs/2606.11680,True,core,execution_state,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,retrieval_representation;execution_multimodal;systems_efficiency,5,agent-memory
+2606.12290,2026-06-10,recent-2026-Apr-Jul,Selection Integrity for LLM Graph Memory: An Accumulability Criterion for Information-Flow-Blind Retrieval,https://arxiv.org/abs/2606.12290,True,core,execution_state,model,retrieval_access;representation_organization;security_privacy_trust;systems_efficiency,retrieval_representation;security_governance,0,agent-memory
+2606.12703,2026-06-10,recent-2026-Apr-Jul,SMSR: Certified Defence Against Runtime Memory Poisoning in Persistent LLM Agent Systems,https://arxiv.org/abs/2606.12703,True,core,execution_state,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,security_governance;systems_efficiency,3,long-term-memory
+2606.12852,2026-06-11,recent-2026-Apr-Jul,WISE: A Long-Horizon Agent in Minecraft with Why-Which Reasoning,https://arxiv.org/abs/2606.12852,False,core,execution_state,model,retrieval_access;representation_organization;execution_state;multimodal_embodied_gui,execution_multimodal,3,episodic-memory
+2606.12945,2026-06-11,recent-2026-Apr-Jul,Learning What to Remember: A Cognitively Grounded Multi-Factor Value Model for Agentic Memory,https://arxiv.org/abs/2606.12945,True,core,execution_state,model,retrieval_access;update_forgetting;evaluation_diagnosis;systems_efficiency,experience_skill_policy,9,agent-memory
+2606.13177,2026-06-11,recent-2026-Apr-Jul,MemRefine: LLM-Guided Compression for Long-Term Agent Memory,https://arxiv.org/abs/2606.13177,True,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal;systems_efficiency,4,agent-memory
+2606.13392,2026-06-11,recent-2026-Apr-Jul,MiniMax Sparse Attention,https://arxiv.org/abs/2606.13392,False,supporting,execution_state,model,retrieval_access;representation_organization;multimodal_embodied_gui;systems_efficiency,,4,long-term-memory
+2606.14470,2026-06-12,recent-2026-Apr-Jul,"GitOfThoughts: Version-Controlled Reasoning and Agent Memory You Can Replay, Diff, and Merge",https://arxiv.org/abs/2606.14470,True,supporting,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,,6,agent-memory
+2606.14571,2026-06-12,recent-2026-Apr-Jul,StreamMemBench: Streaming Evaluation of Agent Memory for Future-Oriented Assistance,https://arxiv.org/abs/2606.14571,True,supporting,execution_state,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis,benchmark_diagnosis,4,agent-memory;memory-evaluation
+2606.15071,2026-06-13,recent-2026-Apr-Jul,Quantum learning with a single-atom sensor,https://arxiv.org/abs/2606.15071,False,supporting,execution_state,model,systems_efficiency,experience_skill_policy,0,agent-memory
+2606.15609,2026-06-14,recent-2026-Apr-Jul,FragFuse: Bypassing Access Control of Large Language Model Agents via Memory-Based Query Fragmentation and Fusion,https://arxiv.org/abs/2606.15609,True,supporting,execution_state,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,,6,agent-memory;long-term-memory
+2606.15903,2026-06-14,recent-2026-Apr-Jul,Control-Plane Placement Shapes Forgetting: An Architectural Study of Agent Memory Across Thirteen System Configurations,https://arxiv.org/abs/2606.15903,True,supporting,execution_state,model,retrieval_access;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,benchmark_diagnosis;update_forgetting,8,agent-memory
+2606.16903,2026-06-15,recent-2026-Apr-Jul,Directory-Aware Query and Maintenance in Vector Databases,https://arxiv.org/abs/2606.16903,False,supporting,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,retrieval_representation,4,agent-memory
+2606.17328,2026-06-15,recent-2026-Apr-Jul,MemTrace: Probing What Final Accuracy Misses in Long-Term Memory,https://arxiv.org/abs/2606.17328,True,supporting,execution_state,model,retrieval_access;execution_state;evaluation_diagnosis;systems_efficiency,,4,long-term-memory
+2606.17628,2026-06-16,recent-2026-Apr-Jul,OPD-Evolver: Cultivating Holistic Agent Evolver via On-Policy Distillation,https://arxiv.org/abs/2606.17628,False,supporting,execution_state,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis,experience_skill_policy,4,agent-memory
+2606.18356,2026-06-16,recent-2026-Apr-Jul,"SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents",https://arxiv.org/abs/2606.18356,False,supporting,execution_state,model,representation_organization;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,,7,long-term-memory
+2606.18406,2026-06-16,recent-2026-Apr-Jul,CoreMem: Riemannian Retrieval and Fisher-Guided Distillation for Long-Term Memory in Dialogue Agents,https://arxiv.org/abs/2606.18406,True,supporting,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,retrieval_representation;experience_skill_policy;execution_multimodal,5,agent-memory
+2606.18746,2026-06-17,recent-2026-Apr-Jul,What Must Generalist Agents Remember?,https://arxiv.org/abs/2606.18746,True,core,experience_learning,model,execution_state,,0,agent-memory
+2606.18829,2026-06-17,recent-2026-Apr-Jul,GateMem: Benchmarking Memory Governance in Multi-Principal Shared-Memory Agents,https://arxiv.org/abs/2606.18829,True,supporting,shared_memory,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis;security_privacy_trust;systems_efficiency,benchmark_diagnosis;security_governance,5,agent-memory;memory-evaluation
+2606.18950,2026-06-17,recent-2026-Apr-Jul,RTSGameBench: An RTS Benchmark for Strategic Reasoning by Vision-Language Models,https://arxiv.org/abs/2606.18950,False,supporting,update_forgetting,model,representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multi_agent_shared,benchmark_diagnosis,3,agent-memory
+2606.19409,2026-06-17,recent-2026-Apr-Jul,OpenRath: Session-Centered Runtime State for Agent Systems,https://arxiv.org/abs/2606.19409,False,core,representation,model,evaluation_diagnosis;security_privacy_trust;multi_agent_shared;systems_efficiency,systems_efficiency,3,agent-memory
+2606.19847,2026-06-18,recent-2026-Apr-Jul,AtomMem: Building Simple and Effective Memory System for LLM Agents via Atomic Facts,https://arxiv.org/abs/2606.19847,True,core,representation,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,,2,long-term-memory
+2606.19857,2026-06-18,recent-2026-Apr-Jul,Large Language Models Do Not Always Need Readable Language,https://arxiv.org/abs/2606.19857,False,supporting,representation,model,representation_organization;evaluation_diagnosis;multi_agent_shared;systems_efficiency,,4,agent-memory
+2606.20047,2026-06-18,recent-2026-Apr-Jul,PACMS: Submodular Context Selection as a Pluggable Engine for LLM Agents,https://arxiv.org/abs/2606.20047,False,supporting,update_forgetting,model,retrieval_access;systems_efficiency,,2,long-term-memory
+2606.20515,2026-06-18,recent-2026-Apr-Jul,S-Agent: Spatial Tool-Use Elicits Reasoning for Spatial Intelligence,https://arxiv.org/abs/2606.20515,False,core,representation,model,representation_organization;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,6,agent-memory
+2606.20954,2026-06-18,recent-2026-Apr-Jul,Learning What Not to Forget: Long-Horizon Agent Memory from a Few Kilobytes of Learning,https://arxiv.org/abs/2606.20954,True,core,representation,model,retrieval_access;update_forgetting;execution_state;evaluation_diagnosis;systems_efficiency,update_forgetting;experience_skill_policy;execution_multimodal,8,agent-memory
+2606.21144,2026-06-19,recent-2026-Apr-Jul,AdaMem: Learning What to Remember for Personalized Long-Horizon LLM Agents,https://arxiv.org/abs/2606.21144,True,core,representation,model,representation_organization;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;systems_efficiency,experience_skill_policy;execution_multimodal;personal_shared,5,long-term-memory
+2606.21562,2026-06-19,recent-2026-Apr-Jul,Compressing Observation History into Agent Memory: Distilling Transformers into Recurrent Transformers,https://arxiv.org/abs/2606.21562,True,supporting,representation,model,representation_organization;execution_state;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;systems_efficiency,2,agent-memory
+2606.21649,2026-06-19,recent-2026-Apr-Jul,EvoEmbedding: Evolvable Representations for Long-Context Retrieval and Agentic Memory,https://arxiv.org/abs/2606.21649,True,supporting,representation,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis,retrieval_representation;experience_skill_policy,6,agent-memory
+2606.22030,2026-06-20,recent-2026-Apr-Jul,Nous: A Predictive World Model for Long-Term Agent Memory,https://arxiv.org/abs/2606.22030,True,core,experience_learning,model,representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,,5,agent-memory;memory-evaluation
+2606.22263,2026-06-20,recent-2026-Apr-Jul,Revelio: Cost-Efficient Agentic Memory Safety Vulnerability Detection For Repository-Scale Codebases,https://arxiv.org/abs/2606.22263,True,supporting,security_privacy,model,evaluation_diagnosis;security_privacy_trust;systems_efficiency,security_governance;systems_efficiency,5,agent-memory
+2606.22844,2026-06-22,recent-2026-Apr-Jul,RaMem: Contextual Reinstatement for Long-term Agentic Memory,https://arxiv.org/abs/2606.22844,True,core,execution_state,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis,,7,agent-memory;long-term-memory;memory-evaluation
+2606.22877,2026-06-22,recent-2026-Apr-Jul,DynamicMem: A Long-Horizon Memory Benchmark in Real-World Settings,https://arxiv.org/abs/2606.22877,True,supporting,security_privacy,model,retrieval_access;update_forgetting;execution_state;evaluation_diagnosis;security_privacy_trust;systems_efficiency,benchmark_diagnosis;execution_multimodal,6,memory-evaluation
+2606.23127,2026-06-22,recent-2026-Apr-Jul,"Managing Procedural Memory in LLM Agents: Control, Adaptation, and Evaluation",https://arxiv.org/abs/2606.23127,True,core,evaluation,model,experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis;experience_skill_policy,7,procedural-memory
+2606.23195,2026-06-22,recent-2026-Apr-Jul,Memory Contagion: Cross-Temporal Propagation of Evaluator Bias via Agent Memory,https://arxiv.org/abs/2606.23195,True,supporting,security_privacy,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,benchmark_diagnosis,4,agent-memory
+2606.23283,2026-06-22,recent-2026-Apr-Jul,Towards Root Memories: Benchmarking and Enhancing Implicit Logical Memory Retrieval for Personalized LLMs,https://arxiv.org/abs/2606.23283,True,core,execution_state,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis;retrieval_representation;personal_shared,5,agent-memory
+2606.24322,2026-06-23,recent-2026-Apr-Jul,"Securing LLM-Agent Long-Term Memory Against Poisoning: Non-Malleable, Origin-Bound Authority with Machine-Checked Guarantees",https://arxiv.org/abs/2606.24322,True,supporting,security_privacy,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,security_governance,6,agent-memory;long-term-memory
+2606.24535,2026-06-23,recent-2026-Apr-Jul,Governed Shared Memory for Multi-Agent LLM Systems,https://arxiv.org/abs/2606.24535,True,core,shared_memory,model,retrieval_access;update_forgetting;evaluation_diagnosis;security_privacy_trust;multi_agent_shared;systems_efficiency,security_governance;personal_shared,7,agent-memory
+2606.24595,2026-06-23,recent-2026-Apr-Jul,MEMPROBE: Probing Long-Term Agent Memory via Hidden User-State Recovery,https://arxiv.org/abs/2606.24595,True,supporting,security_privacy,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,5,agent-memory;long-term-memory
+2606.24775,2026-06-23,recent-2026-Apr-Jul,Are We Ready For An Agent-Native Memory System?,https://arxiv.org/abs/2606.24775,True,core,update_forgetting,model,retrieval_access;representation_organization;update_forgetting;execution_state;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,6,agent-memory
+2606.25115,2026-06-23,recent-2026-Apr-Jul,Forget to Improve: On-Device LLM-Agent Continual Learning via Budget-Curated Memory,https://arxiv.org/abs/2606.25115,True,supporting,security_privacy,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui;systems_efficiency,update_forgetting;experience_skill_policy;systems_efficiency,2,agent-memory
+2606.25161,2026-06-23,recent-2026-Apr-Jul,TRUSTMEM: Learning Trustworthy Memory Consolidation for LLM Agents with Long-Term Memory,https://arxiv.org/abs/2606.25161,True,core,update_forgetting,model,update_forgetting;experience_skill_learning;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,update_forgetting;experience_skill_policy;security_governance,7,agent-memory;long-term-memory
+2606.25206,2026-06-23,recent-2026-Apr-Jul,RAVEN: Long-Horizon Reasoning & Navigation with a Visuo-Spatio-Temporal Memory,https://arxiv.org/abs/2606.25206,True,core,execution_state,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,execution_multimodal,2,agent-memory
+2606.26627,2026-06-25,recent-2026-Apr-Jul,Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents,https://arxiv.org/abs/2606.26627,False,supporting,security_privacy,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust,benchmark_diagnosis;security_governance,2,agent-memory
+2606.26790,2026-06-25,recent-2026-Apr-Jul,OPID: On-Policy Skill Distillation for Agentic Reinforcement Learning,https://arxiv.org/abs/2606.26790,False,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,6,procedural-memory
+2606.27499,2026-06-25,recent-2026-Apr-Jul,DMV-Bench: Diagnosing Long-Horizon Multimodal Agents' Visual Memory with Incidental Cue Injection,https://arxiv.org/abs/2606.27499,True,core,execution_state,model,retrieval_access;execution_state;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,benchmark_diagnosis;execution_multimodal,6,agent-memory
+2606.28781,2026-06-27,recent-2026-Apr-Jul,HyphaeDB: A Living Knowledge Topology for Agent-First Memory,https://arxiv.org/abs/2606.28781,True,core,execution_state,model,retrieval_access;representation_organization;multi_agent_shared;systems_efficiency,systems_efficiency,0,agent-memory
+2606.28971,2026-06-27,recent-2026-Apr-Jul,Self-Evolving Agentic Image Restoration via Deliberate Planning and Intuitive Execution,https://arxiv.org/abs/2606.28971,False,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,3,episodic-memory
+2606.29774,2026-06-29,recent-2026-Apr-Jul,Analytic Concept-Centric Memory for Agentic Embodied Manipulation,https://arxiv.org/abs/2606.29774,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,5,agent-memory;memory-evaluation
+2606.29778,2026-06-29,recent-2026-Apr-Jul,Mandol: An Agglomerative Agent Memory System for Long-Term Conversations,https://arxiv.org/abs/2606.29778,True,core,execution_state,model,retrieval_access;representation_organization;update_forgetting;evaluation_diagnosis;systems_efficiency,,4,agent-memory
+2606.29788,2026-06-29,recent-2026-Apr-Jul,MemLeak: Diagnosing Information Leaks in Multimodal Agent Memory,https://arxiv.org/abs/2606.29788,True,core,update_forgetting,model,representation_organization;update_forgetting;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,benchmark_diagnosis;execution_multimodal;security_governance,7,agent-memory
+2606.29824,2026-06-29,recent-2026-Apr-Jul,Neural Procedural Memory: Empowering LLM Agents with Implicit Activation Steering,https://arxiv.org/abs/2606.29824,True,core,update_forgetting,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy,5,agent-memory;procedural-memory
+2606.29914,2026-06-29,recent-2026-Apr-Jul,MemDelta: Controlled Baselines and Hidden Confounds in Agent Memory Evaluation,https://arxiv.org/abs/2606.29914,True,supporting,security_privacy,model,retrieval_access;update_forgetting;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis,7,agent-memory;memory-evaluation
+2606.29961,2026-06-29,recent-2026-Apr-Jul,DuoMem: Towards Capable On-Device Memory Agents via Dual-Space Distillation,https://arxiv.org/abs/2606.29961,True,core,update_forgetting,model,experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui,experience_skill_policy;systems_efficiency,5,agent-memory
+2606.30296,2026-06-29,recent-2026-Apr-Jul,ManimAgent: Self-Evolving Multimodal Agents for Visual Education,https://arxiv.org/abs/2606.30296,False,core,execution_state,model,retrieval_access;update_forgetting;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy;execution_multimodal,6,episodic-memory
+2606.30566,2026-06-29,recent-2026-Apr-Jul,Forensic Trajectory Signatures for Agent Memory Poisoning Detection,https://arxiv.org/abs/2606.30566,True,core,update_forgetting,model,retrieval_access;execution_state;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,security_governance,4,agent-memory;long-term-memory
+2606.30639,2026-06-29,recent-2026-Apr-Jul,Self-Evolving World Models for LLM Agent Planning,https://arxiv.org/abs/2606.30639,False,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis,experience_skill_policy,4,episodic-memory
+2606.31046,2026-06-30,recent-2026-Apr-Jul,OpenLife: Toward Open-World Artificial Life with Autonomous LLM Agents,https://arxiv.org/abs/2606.31046,False,core,experience_learning,model,retrieval_access;representation_organization;experience_skill_learning;evaluation_diagnosis;systems_efficiency,,2,long-term-memory
+2606.31612,2026-06-30,recent-2026-Apr-Jul,What Memory Do GUI Agents Really Need? From Passive Records to Active Task-Driving States,https://arxiv.org/abs/2606.31612,True,core,update_forgetting,model,retrieval_access;update_forgetting;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,benchmark_diagnosis;execution_multimodal,3,agent-memory
+2607.02579,2026-06-30,recent-2026-Apr-Jul,When Not to Write Memory: Governing False Promotion from Correlated Agent Traces,https://arxiv.org/abs/2607.02579,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,update_forgetting;security_governance,4,agent-memory
+2607.00454,2026-07-01,recent-2026-Apr-Jul,Agri-SAGE: Simulation-Grounded Multi-Agent LLM for Context-Aware Agricultural Advisory Generation,https://arxiv.org/abs/2607.00454,False,core,update_forgetting,model,retrieval_access;evaluation_diagnosis;multi_agent_shared;multimodal_embodied_gui;systems_efficiency,personal_shared,6,episodic-memory
+2607.01047,2026-07-01,recent-2026-Apr-Jul,Conversable Complexity: Agentic LLM Collectives as Interpretable Substrates,https://arxiv.org/abs/2607.01047,False,core,execution_state,model,retrieval_access;experience_skill_learning;evaluation_diagnosis,personal_shared,3,long-term-memory
+2607.01071,2026-07-01,recent-2026-Apr-Jul,MemSyco-Bench: Benchmarking Sycophancy in Agent Memory,https://arxiv.org/abs/2607.01071,True,core,execution_state,model,retrieval_access;update_forgetting;evaluation_diagnosis;systems_efficiency,benchmark_diagnosis,4,agent-memory;memory-evaluation
+2607.01523,2026-07-01,recent-2026-Apr-Jul,Multi-Head Recurrent Memory Agents,https://arxiv.org/abs/2607.01523,True,core,update_forgetting,model,update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,7,agent-memory
+2607.01709,2026-07-02,recent-2026-Apr-Jul,COMFYCLAW: Self-Evolving Skill Harnesses for Image Generation Workflows,https://arxiv.org/abs/2607.01709,False,core,update_forgetting,model,retrieval_access;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,experience_skill_policy,5,agent-memory
+2607.01916,2026-07-02,recent-2026-Apr-Jul,ContextSniper: AntTrail's Token-Efficient Code Memory for Repository-Level Program Repair,https://arxiv.org/abs/2607.01916,True,core,update_forgetting,model,retrieval_access;representation_organization;evaluation_diagnosis;systems_efficiency,systems_efficiency,6,agent-memory
+2607.01935,2026-07-02,recent-2026-Apr-Jul,A-TMA: Decoupling State-Aware Memory Failures in Long-Term Agent Memory,https://arxiv.org/abs/2607.01935,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis,,3,agent-memory;long-term-memory
+2607.03726,2026-07-04,recent-2026-Apr-Jul,SelfMem: Self-Optimizing Memory for AI Agents,https://arxiv.org/abs/2607.03726,True,core,update_forgetting,model,retrieval_access;execution_state;experience_skill_learning;evaluation_diagnosis;multimodal_embodied_gui;systems_efficiency,,7,agent-memory
+2607.04089,2026-07-05,recent-2026-Apr-Jul,PLACEMEM: Toward a Compute-Aware Memory Plane for Lifelong Agents,https://arxiv.org/abs/2607.04089,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;security_privacy_trust;systems_efficiency,,2,agent-memory
+2607.04391,2026-07-05,recent-2026-Apr-Jul,Memory-Orchestrated Semantic System (MOSS): An Auditable Agentic Memory Architecture,https://arxiv.org/abs/2607.04391,True,core,update_forgetting,model,retrieval_access;representation_organization;systems_efficiency,,0,agent-memory
+2607.05029,2026-07-06,recent-2026-Apr-Jul,Your Agent's Memories Are Not Its Own: Forged Reasoning Attacks on LLM Agent Memory and Defenses,https://arxiv.org/abs/2607.05029,True,core,update_forgetting,model,retrieval_access;evaluation_diagnosis;security_privacy_trust,security_governance,6,agent-memory;long-term-memory
+2607.05577,2026-07-06,recent-2026-Apr-Jul,Narrative World Model: Narratology-Grounded Writer Memory for Long-Form Fiction,https://arxiv.org/abs/2607.05577,True,core,update_forgetting,model,retrieval_access;representation_organization;evaluation_diagnosis,,5,agent-memory
+2607.06595,2026-07-06,recent-2026-Apr-Jul,When Agents Remember Too Much: Memory Poisoning Attacks on Large Language Model Agents,https://arxiv.org/abs/2607.06595,True,core,update_forgetting,model,retrieval_access;execution_state;security_privacy_trust;systems_efficiency,security_governance,4,agent-memory;long-term-memory
+2607.06195,2026-07-07,recent-2026-Apr-Jul,LogicHunter: Testing LLM Agent Frameworks with an Agentic Oracle,https://arxiv.org/abs/2607.06195,False,core,update_forgetting,model,retrieval_access;representation_organization;evaluation_diagnosis;security_privacy_trust;multimodal_embodied_gui,,7,memory-evaluation
+2607.07108,2026-07-08,recent-2026-Apr-Jul,Seeing and Reflecting: Multimodal Memory-Enhanced Agent Collaboration for Recommendation,https://arxiv.org/abs/2607.07108,True,core,update_forgetting,model,representation_organization;update_forgetting;experience_skill_learning;multimodal_embodied_gui,execution_multimodal,5,agent-memory
+2607.08032,2026-07-09,recent-2026-Apr-Jul,"What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents",https://arxiv.org/abs/2607.08032,True,core,update_forgetting,model,retrieval_access;update_forgetting;evaluation_diagnosis;systems_efficiency,update_forgetting,5,agent-memory
+2607.08716,2026-07-09,recent-2026-Apr-Jul,Remember When It Matters: Proactive Memory Agent for Long-Horizon Agents,https://arxiv.org/abs/2607.08716,True,core,update_forgetting,model,retrieval_access;representation_organization;execution_state;evaluation_diagnosis;multimodal_embodied_gui,execution_multimodal,6,agent-memory