Newest papers

Recoverability and Verifier Strength: Two Measured Preconditions for Unattended Agent Self-Improvement

Wang De, Peng Qi · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.27465 · cited by 0 works (OpenAlex) · open-access PDF

Korean

에이전트 실행을 실행하는 동안이 에이전트는 자신의 에이전트를 개선하기 시작합니다 : 그들은 배포 된 에이전트의 실행 흔적을 읽고, 변경을 생성하고, 자동 검증자가 그것을 승인하고 인간 작가가없는 작업을 해제합니다. 그래서이 루프의 얼마나 안전하게 인간없이 실행할 수 있습니까? 우리는 플랫폼을 측정 할 수있는 두 개의 양을 제안합니다. 검증 강도 $σ$는 사용할 수있는 롤백 행동의 얼마나 잘 복원됩니다. LLM 판사가 종종 변경을 수행 할 때 - 신뢰할 수있는 참조 검사자와 동의합니다. 그것은 신속하게 작업을 수행 할 수있는 얼마나 많은 단계를 제한합니다. 복구 가능성은 나중에 나쁜 변경이

Machine translation (Meta M2M100); the English original follows.

Abstract

Agent platforms are starting to improve their own agents: they read a deployed agent's execution traces, derive a change, have an automated verifier approve it, and release it with no human author. How much of this loop can safely run without a human? We propose two quantities a platform can measure. Verifier strength $σ$ is how well the available verifier -- often an LLM judge -- agrees with a trusted reference verifier; it limits how much a step can improve the agent. Recoverability is how cheaply a bad change can be undone: whether undoing it needs a new verifier approval, and how much of the old behavior a rollback restores. A step should run unattended only if both are adequate. On our platform this rule allows prompt, tool and skill edits to run unattended and keeps weight updates under human sign-off. Our main contribution is these two definitions and how to measure them. We built a platform that applies the rule and tested it in two ways: on the platform's own approval gate, and in controlled experiments on the $τ^2$-bench agent benchmark. On verifier strength, the platform's LLM judge sometimes gave the same score no matter what it was shown (9 of 37 settings) and reported no error, so a verifier can look healthy while approving anything; its strength must be measured, and two simple tests catch the failure. On recoverability, restoring a saved copy brought the agent back to its original behavior, but a regenerated undo did not, so only changes whose old version is kept can be reversed without a human; and while a change was deployed, it altered the outcome of 17 percent of tasks beyond normal run-to-run variation, which no undo repairs. Together these support the rule: a change can run unattended only when its old version is kept and its verifier's strength has been measured.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Undo, Undoing, Computer science, Simple (philosophy), Computer security, Measure (data warehouse)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange