按年被提到
20232023 · 1 次2026
大模型与语言模型›大模型推理与思维链·arXiv:2305.13160 · cs.CL

Can ChatGPT Defend its Belief in Truth? Evaluating LLM Reasoning via Debate

Boshi Wang, Xiang Yue, Huan Sun · 2023
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 2 人首次 2023-12-08 · 最近 2024-01-11
2不同的人
2活跃月份
2讨论串
71被引(OpenAlex)
arXiv 摘要(候选层不写导读)

Large language models (LLMs) such as ChatGPT and GPT-4 have shown impressive performance in complex reasoning tasks. However, it is difficult to know whether the models are reasoning based on deep understandings of truth and logic, or leveraging their memorized patterns in a relatively superficial way. In this work, we explore testing LLMs' reasoning by engaging with them in a debate-like conversation, where given a question, the LLM and the user need to discuss to make the correct decision starting from opposing arguments. Upon mitigating the Clever Hans effect, our task requires the LLM to not only achieve the correct answer on its own, but also be able to hold and defend its belief instead of blindly believing or getting misled by the user's (invalid) arguments and critiques, thus testing in greater depth whether the LLM grasps the essence of the reasoning required to solve the problem. Across a range of complex reasoning benchmarks spanning math, commonsense, logic and BIG-Bench tasks, we find that despite their impressive performance as reported in existing work on generating correct step-by-step solutions in the beginning, LLMs like ChatGPT cannot maintain their beliefs in truth for a significant portion of examples when challenged by oftentimes absurdly invalid arguments. Our work points to danger zones of model alignment, and also suggests more careful treatments and interpretations of the recent findings that LLMs can improve their responses based on feedback.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论Paul-Craft · 2024-01

“Indeed, ChatGPT will frequently backpedal if you even slightly hint that its answer may have been wrong: https://arxiv.org/pdf/2305.13160.pdf I have found that you can reduce that behavior if you prompt it correctly, though.”

How to get coworkers to stop giving me ChatGPT-generated suggestions?原评论 ↗
被提交为 story · 1 次