跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2603.01608 · v3

Evaluating and Understanding Scheming Propensity in LLM Agents

首次提交 2026-03-02 16:38
版本更新 2026-11-17 12:32
雷达观测 2026-11-18 19:43
arXiv ↗PDF ↗

摘要

As frontier language models are increasingly deployed as autonomous agents pursuing complex, long-term objectives, there is increased risk of scheming: agents covertly pursuing misaligned goals. Prior work has focused on showing agents are capable of scheming, but their propensity to scheme in realistic scenarios remains underexplored. To understand when agents scheme, we decompose scheming incentives into agent factors and environmental factors. We develop realistic settings allowing us to systematically vary these factors, each with scheming opportunities for agents that pursue instrumentally convergent goals such as self-preservation, resource acquisition, and goal-guarding. Across 11 frontier models and four scenarios, we find almost no scheming despite high environmental incentives, and show this is unlikely to be explained by verbalized evaluation awareness. While inserting adversarially-designed prompt snippets that encourage agency and goal-directedness into an agent's system prompt can induce high scheming rates, snippets used in real agent scaffolds rarely do. In model organisms built with these snippets, scheming behavior is brittle: removing a single tool can drop the scheming rate from 59% to 3%. Increasing oversight can raise rather than deter scheming by up to 25%, an effect that depends on the wording of the oversight sentence and does not appear in a second scenario, where the sabotage itself is reviewed. These results suggest that scheming propensity should be reported per agent configuration rather than as a single rate. Our incentive decomposition enables such systematic measurement in settings relevant for deployment, which is necessary as agents are entrusted with increasingly consequential tasks.

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

本站首次收录为 v3,暂无更早版本可对照。

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。

v3 · 2026-11-17 12:32 · 最新观测版本

Evaluating and Understanding Scheming Propensity in LLM Agents

As frontier language models are increasingly deployed as autonomous agents pursuing complex, long-term objectives, there is increased risk of scheming: agents covertly pursuing misaligned goals. Prior work has focused on showing agents are capable of scheming, but their propensity to scheme in realistic scenarios remains underexplored. To understand when agents scheme, we decompose scheming incentives into agent factors and environmental factors. We develop realistic settings allowing us to systematically vary these factors, each with scheming opportunities for agents that pursue instrumentally convergent goals such as self-preservation, resource acquisition, and goal-guarding. Across 11 frontier models and four scenarios, we find almost no scheming despite high environmental incentives, and show this is unlikely to be explained by verbalized evaluation awareness. While inserting adversarially-designed prompt snippets that encourage agency and goal-directedness into an agent's system prompt can induce high scheming rates, snippets used in real agent scaffolds rarely do. In model organisms built with these snippets, scheming behavior is brittle: removing a single tool can drop the scheming rate from 59% to 3%. Increasing oversight can raise rather than deter scheming by up to 25%, an effect that depends on the wording of the oversight sentence and does not appear in a second scenario, where the sabotage itself is reviewed. These results suggest that scheming propensity should be reported per agent configuration rather than as a single rate. Our incentive decomposition enables such systematic measurement in settings relevant for deployment, which is necessary as agents are entrusted with increasingly consequential tasks.

查看此版本解读 →查看此版本原文 ↗