跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2602.17544 · v2

Evaluating Chain-of-Thought Reasoning through Reusability and Verifiability

首次提交 2026-02-20 00:59
版本更新 2026-11-13 22:57
雷达观测 2026-11-16 11:41
arXiv ↗PDF ↗

摘要

AI applications increasingly compose multiple models, where one model acts on the output produced by another model. The transferred text can also carry the reasoning of the sender model. Chain-of-thought is our experimental instance of such text. We ask three questions about such inter-model reasoning transfer. Is the reasoning transmitted reliably? Is it useful to the receiver? Is it safe? We study these three questions with a Sender-Receiver setup. A Sender LLM produces a reasoning trace. A separate Receiver LLM must act on it. Neither model knows the identity of the other. The Sender therefore cannot tailor its reasoning to a particular Receiver. Reliability asks whether the Receiver recovers the Sender's intended conclusion. We measure it with Verifiability. Usefulness asks whether transferred reasoning helps a Receiver correct its mistakes. We measure it with Correctability. Safety asks whether transferred reasoning makes a Receiver abandon a correct answer and follow the wrong reasoning supplied by the Sender. We measure it with Deceptiveness. All three factors (Task, Sender, and Receiver) affect all three metrics that we measure. However, there are some interesting trends. Nature of task is a more important factor in the reliability of reasoning transfer. Reliability is higher on arithmetic tasks and lower on science and commonsense tasks. The nature of Receiver plays more important role while deciding the usefulness of reasoning transfer. The nature of the Sender matters most for whether corrupted reasoning harms the Receiver.

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 v2,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。

v2 · 2026-11-13 22:57 · 最新观测版本

Evaluating Chain-of-Thought Reasoning through Reusability and Verifiability

AI applications increasingly compose multiple models, where one model acts on the output produced by another model. The transferred text can also carry the reasoning of the sender model. Chain-of-thought is our experimental instance of such text. We ask three questions about such inter-model reasoning transfer. Is the reasoning transmitted reliably? Is it useful to the receiver? Is it safe? We study these three questions with a Sender-Receiver setup. A Sender LLM produces a reasoning trace. A separate Receiver LLM must act on it. Neither model knows the identity of the other. The Sender therefore cannot tailor its reasoning to a particular Receiver. Reliability asks whether the Receiver recovers the Sender's intended conclusion. We measure it with Verifiability. Usefulness asks whether transferred reasoning helps a Receiver correct its mistakes. We measure it with Correctability. Safety asks whether transferred reasoning makes a Receiver abandon a correct answer and follow the wrong reasoning supplied by the Sender. We measure it with Deceptiveness. All three factors (Task, Sender, and Receiver) affect all three metrics that we measure. However, there are some interesting trends. Nature of task is a more important factor in the reliability of reasoning transfer. Reliability is higher on arithmetic tasks and lower on science and commonsense tasks. The nature of Receiver plays more important role while deciding the usefulness of reasoning transfer. The nature of the Sender matters most for whether corrupted reasoning harms the Receiver.

查看此版本解读 →查看此版本原文 ↗