feat: add reduced AttnRes trace lab

This commit is contained in:
wuyang
2026-07-30 07:30:01 +08:00
parent d1d9d22bf3
commit 4ce780dcc9
11 changed files with 1070 additions and 26 deletions
+9 -9
View File
@@ -107,7 +107,7 @@ const paths = [
</div>
<aside class="hero-aside" aria-label="项目统计">
<span>RESEARCH CUTOFF</span>
<strong>29 · 07 · 2026</strong>
<strong>30 · 07 · 2026</strong>
<p>首版持续建设中,所有动态结论带日期</p>
<div class="hero-stats">
<div><b>17</b><span>核心专题</span></div>
@@ -128,20 +128,20 @@ const paths = [
<div class="release-grid">
<a class="release-card k3-release" href="/k3/">
<div>
<p class="eyebrow"><span>NEW / K3 ROUND 03</span> REPORT · CHECKPOINT · KERNEL · BOUNDARY</p>
<h2>47 页不再压成摘要:再把 1.56 TB 开放工件接回报告</h2>
<p class="eyebrow"><span>NEW / K3 ROUND 04</span> ATTENTION RESIDUALS · PREREGISTERED PROBE</p>
<h2>真实 K3 权重仍有冲突:先把一个可证伪的深度路由问题完整做完</h2>
<p>
在三十二张报告问题账之外,继续审计 96 个 safetensors 分片、497,220 个 tensor entries、
真实 KDA / MLA / MoE / MoonViT shape、小范围权重统计,并把 FlashKDA 推进到 RTX 5090
6/6 exact-match、K3 fixed / varlen 计时与未决 checkpoint 形状矛盾。
不裁剪 <code>A_log [128]</code> 冒充 96-head K3 forward;冻结相同主干、初始化、数据窗口和预算,
从零训练 Baseline / Full / Block 共 9 个 2,000-step 格。两种 AttnRes 的三 seed BPC
配对方向都为负,但梯度均匀性指标没有复现论文叙述;支持与反证在同一实验室展示。
</p>
</div>
<dl>
<div><dt>REPORT</dt><dd>16 Figures · 5 Tables</dd></div>
<div><dt>ARTIFACTS</dt><dd>96 shards · 497,220 entries</dd></div>
<div><dt>KERNEL</dt><dd>sm_120a · exact 6/6</dd></div>
<div><dt>GRID</dt><dd>3 structures × 3 seeds</dd></div>
<div><dt>REPLAY</dt><dd>2,000 steps · 8/8 exact</dd></div>
</dl>
<span class="release-arrow" aria-hidden="true">从报告目录进入开放工件证据链 →</span>
<span class="release-arrow" aria-hidden="true">进入训练曲线、深度权重与梯度反证 →</span>
</a>
<a class="release-card deepseek-release" href="/deepseek/">
<div>