跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.21946 · v1

Conserved Verdicts, Drifting Commitments: Identifying Harmful Overthinking

首次提交 2026-11-17 06:08
版本更新 2026-11-17 06:08
雷达观测 2026-11-18 19:43
arXiv ↗PDF ↗

摘要

A reasoning model that has reached a correct answer can lose it by continuing to think. Deciding whether to stop means comparing two probabilities at the same point of a chain: that the answer the model would give if stopped now is correct (its commitment), and that the answer it will finish with is correct (its verdict). We show that the verdict alone, which is what final-correctness probes estimate, cannot make this comparison. As a forecast of a fixed outcome it has no predictable drift, whether continuing helps or hurts. The commitment-verdict gap equals the expected remaining change in answer quality and the accuracy gained by stopping. A second identity splits any accuracy-versus-budget curve into answer change within chains and change in which chains are still running. We measure both on MATH-500. At all 379 points where DeepSeek-R1-Distill-Qwen-1.5B holds a correct answer, selected without regard to how its chain ends, continuing loses 3.6 points on average (95% CI [2.2, 5.2]), and the loss concentrates in a few problems. In six reasoning models from three families, on MATH-500 and OlympiadBench and with budgets up to 16,384 tokens, declining accuracy curves come from which chains remain, while the chains still reasoning keep or improve their answers. An oracle stopping rule gains at most +1.8 and +2.5 points at 1.5B and 7B. The framework says what to measure before building a stopping rule.

作者备注:40 pages

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 · 2026-11-17 06:08 · 最新观测版本

Conserved Verdicts, Drifting Commitments: Identifying Harmful Overthinking

A reasoning model that has reached a correct answer can lose it by continuing to think. Deciding whether to stop means comparing two probabilities at the same point of a chain: that the answer the model would give if stopped now is correct (its commitment), and that the answer it will finish with is correct (its verdict). We show that the verdict alone, which is what final-correctness probes estimate, cannot make this comparison. As a forecast of a fixed outcome it has no predictable drift, whether continuing helps or hurts. The commitment-verdict gap equals the expected remaining change in answer quality and the accuracy gained by stopping. A second identity splits any accuracy-versus-budget curve into answer change within chains and change in which chains are still running. We measure both on MATH-500. At all 379 points where DeepSeek-R1-Distill-Qwen-1.5B holds a correct answer, selected without regard to how its chain ends, continuing loses 3.6 points on average (95% CI [2.2, 5.2]), and the loss concentrates in a few problems. In six reasoning models from three families, on MATH-500 and OlympiadBench and with budgets up to 16,384 tokens, declining accuracy curves come from which chains remain, while the chains still reasoning keep or improve their answers. An oracle stopping rule gains at most +1.8 and +2.5 points at 1.5B and 7B. The framework says what to measure before building a stopping rule.

查看此版本解读 →查看此版本原文 ↗