跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.24491 · v1

Probing agent overeagerness in realistic settings with continuation audits

首次提交 2026-11-18 22:23
版本更新 2026-11-18 22:23
雷达观测 2026-11-19 15:04
arXiv ↗PDF ↗

摘要

Frontier coding agents increasingly exhibit overeagerness, taking surprising, sometimes reckless actions beyond a user's mandate --- reported in recent system cards from frontier model developers. These cases are surfaced through internal deployment monitoring, which requires private traffic and is out of reach for open-weight models and independent evaluators. Automated alignment audits are the natural alternative, but capable models increasingly recognise the synthetic setup and adjust their behaviour. We introduce Continuation Audits, a pipeline that grounds the auditor on a prefix drawn from a real coding-agent transcript and directs it to continue the trajectory toward a seeded behaviour. Because the prefix is drawn from a real session, the auditor only needs to maintain a realistic setting rather than manufacturing it. We validate the method on SWE-Chat prefixes: on a paired subset of 90 audits a judge cannot reliably distinguish continuations from their source sessions (realism win rate 0.42 vs.\ 0.43), and across all 900 audits verbalized eval-awareness does not exceed the rate observed in the unaudited real sessions. Applying the pipeline to three recent models across three seeds, we find each model leads a different seed --- DeepSeek~V4~Pro on colleague-credentials misuse, GPT-5.6~Terra on file-permission bypasses, and Kimi~K3 on treating diagnostic questions as instructions. Kimi~K3 additionally continues to bypass file-permission protections after an explicit CLAUDE.md prohibition while the other two stop entirely, corroborating Moonshot's own report of excessive proactiveness.

作者备注:Presented at NeurIPS Workshops: (1) Third Workshop on Agents in the Wild: Safety, Security, and Beyond, and (2) The 1st Workshop on Interpreting Agent Behavior (IAB) at NeurIPS 2026

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

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 · 2026-11-18 22:23 · 最新观测版本

Probing agent overeagerness in realistic settings with continuation audits

Frontier coding agents increasingly exhibit overeagerness, taking surprising, sometimes reckless actions beyond a user's mandate --- reported in recent system cards from frontier model developers. These cases are surfaced through internal deployment monitoring, which requires private traffic and is out of reach for open-weight models and independent evaluators. Automated alignment audits are the natural alternative, but capable models increasingly recognise the synthetic setup and adjust their behaviour. We introduce Continuation Audits, a pipeline that grounds the auditor on a prefix drawn from a real coding-agent transcript and directs it to continue the trajectory toward a seeded behaviour. Because the prefix is drawn from a real session, the auditor only needs to maintain a realistic setting rather than manufacturing it. We validate the method on SWE-Chat prefixes: on a paired subset of 90 audits a judge cannot reliably distinguish continuations from their source sessions (realism win rate 0.42 vs.\ 0.43), and across all 900 audits verbalized eval-awareness does not exceed the rate observed in the unaudited real sessions. Applying the pipeline to three recent models across three seeds, we find each model leads a different seed --- DeepSeek~V4~Pro on colleague-credentials misuse, GPT-5.6~Terra on file-permission bypasses, and Kimi~K3 on treating diagnostic questions as instructions. Kimi~K3 additionally continues to bypass file-permission protections after an explicit CLAUDE.md prohibition while the other two stop entirely, corroborating Moonshot's own report of excessive proactiveness.

查看此版本解读 →查看此版本原文 ↗