agent/papers/items/2026-2605-25310-tool-call-dependency-structure-is-linearly-decodable-in-llm-agent-residual-strea.md

Paper: Tool-Call Dependency Structure is Linearly Decodable in LLM Agent Residual Streams

---

type: paper title: Tool-Call Dependency Structure is Linearly Decodable in LLM Agent Residual Streams authors: Tianda Sun, Dimitar Kazakov year: 2026 venue: arXiv url: https://arxiv.org/abs/2605.25310 code_url: source: arxiv collected_at: 2026-07-08 published_at: 2026-05-25 updated_at: 2026-05-25 status: queued relevance: high topics:

• agent-evaluation
• planning
• tool-use methods:

benchmarks:

models:

datasets:

• cs.CL related_concepts:

related_jobs:

related_experiments:

related_projects:

collection_score: 14 collection_queries: planning-agent

One-line Takeaway

Auto-collected from arXiv because it matched the Agent collection queries. Needs human skim.

Why Collected

• matched queries: planning-agent
• inferred topics: agent-evaluation, planning, tool-use
• arXiv categories: cs.CL
• collection score: 14

Review Checklist

• Does this paper directly inform Agent architecture, evaluation, memory, tools, safety, coding agents, GUI/browser agents, or multi-agent workflows?
• Does it include a benchmark, dataset, code, or reproducible experimental setup?
• Should it be promoted from queued to skimmed or summarized?

Links

• arXiv: https://arxiv.org/abs/2605.25310