Findings of the Association for Computational Linguistics: ACL 2025 / 奖励驱动进化

Confidence Improves Self-Consistency in LLMs

作者：Amir Taubenfeld, Tom Sheffer, Eran Ofek, Amir Feder, Ariel Goldstein, Zorik Gekhman, Gal Yona · 单位：Google Research, The Hebrew University of Jerusalem, Columbia University · 会议/期刊：Findings of the Association for Computational Linguistics: ACL 2025 · 方向：奖励驱动进化 · 发布日期：2025-02-10

自我进化智能体奖励驱动进化大语言模型智能体奖励学习强化学习
新标签打开 AlphaXiv新标签打开 Code
Confidence Improves Self-Consistency in LLMs figure
AlphaXiv 中文概览（可滚动查看）