← 研踪 PaperTrailarXiv:2611.06563 · v1
EPISTEME: Trustworthy Coding Agents Through Tool Call Intent
首次提交 2026-11-04 22:26
版本更新 2026-11-04 22:26
雷达观测 2026-11-05 12:13
摘要
Coding agents change software through tool calls. Tool calls mutate the local state. For good? For bad? Nobody knows. Agents never say why a tool call is made. We propose the epistemic agent design, in which the harness guarantees that the model declares why a tool call is made. The harness then records what the command actually did and checks the self-reported declaration against these effects. The result is a journal 1) that can guard calls before they run and 2) that supports audits and forensics. For evaluation, we use ImpossibleBench, whose tasks can be passed only by cheating, and demonstrate that forcing the agents to say why results in exposed cheating. Across 49 reward-hacking runs of two open-weight models, the journal exposed 90%.
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-04 22:26 · 最新观测版本
EPISTEME: Trustworthy Coding Agents Through Tool Call Intent
Coding agents change software through tool calls. Tool calls mutate the local state. For good? For bad? Nobody knows. Agents never say why a tool call is made. We propose the epistemic agent design, in which the harness guarantees that the model declares why a tool call is made. The harness then records what the command actually did and checks the self-reported declaration against these effects. The result is a journal 1) that can guard calls before they run and 2) that supports audits and forensics. For evaluation, we use ImpossibleBench, whose tasks can be passed only by cheating, and demonstrate that forcing the agents to say why results in exposed cheating. Across 49 reward-hacking runs of two open-weight models, the journal exposed 90%.
查看此版本解读 →查看此版本原文 ↗