aifollow.news
返回 Rohan Paul
Rohan Paul· @rohanpaul_ai · X· · 原发布时间 AI 评分54

微软论文提出“自我分歧”训练,让模型面对错误质疑时更能守住正确答案

AI 导读

研究关注模型在用户或其他智能体提出错误反驳时放弃正确答案的问题。“自我分歧”训练让模型先作答、再提出反对理由并重新裁决;在错误质疑下,两款受测模型保留正确答案的比例从53.3%和25.1%升至约89%。这些结果来自所述模型和测试条件。

原帖引用 · 原文

该语言的正文暂不可用,当前显示已有版本。

@rohanpaul_ai

– https://t.co/a8WoNs5mbD

Title: "Benchmarks Don't Talk Back: Learning Through Self-Disagreement"

回复Rohan Paul@rohanpaul_ai
New Microsoft paper finds that LLMs give up correct answers when challenged with wrong arguments, and accuracy training does not fix it. When a user or another agent disputes an answer, models often abandon correct answers even when the challenge is wrong. More accuracy training, more sampling, or self-reflection does not fix it. Self-Disagreement trains one model to state an answer, write the strongest case against it, then rule on that case and answer again. A standard answer checker rewards the final answer and the ruling. When challengers were wrong, base Qwen3-8B kept its correct answer 53.3% of the time and Gemma-2 9B 25.1%. After training, both kept it about 89% of the time, while accuracy-only training added no such skill. If your model faces pushback from users or other agents, test how often it drops correct answers, since accuracy scores will not show it.
在 X 查看这条帖子 ↗

来源:Rohan Paul · x.com

论文
发现内容有误?提交纠错