CaRL：训练大语言模型在无效推理时及时放弃

AI 摘要

CaRL是一种通过强化学习和事后拒绝增强来减少大语言模型无效推理的方法，同时保持任务性能。研究发现模型在超出能力范围的任务上会产生看似合理但包含细微错误的推理，CaRL通过奖励塑造和拒绝监督来对齐模型行为与能力边界。实验表明该方法能显著减少无效推理，且不牺牲任务性能。

正文节选

Knowing When to Quit: Diagnosing and Training LLMs to Abort Futile Reasoning Abstract CaRL uses reinforcement learning with refusal incentives and hindsight augmentation to reduce futile reasoning in large language models while preserving task performance. Large language models generate computationally expensive yet semantically void reasoning on beyond-capability tasks, creating risks where plausible-sounding but incorrect derivations mislead users. We characterize this futile reasoning phenome

发布时间：—

抓取时间：2026-08-14 15:51

来源机构：Hugging Face