让大模型学会角色认知，通过显式暴露知识盲区提升推理能力
PICTURE: Enhancing Theory-of-Mind in Large Language Models by Revealing, Not Hiding, Characters' Lack of Knowledge
- 不隐藏未知事件，改用自由格式思维链暴露角色知识盲区
- 在虚假信念任务上平均提升7.3%准确率，优于现有方法
- 适合研究心智理论、对话系统与推理增强的学者和开发者
模拟人类心智理论（ToM）一直是自然语言处理中的长期难题。现有方法采用事件隐藏（即视角代入）的推理步骤，但在回答问题前移除角色不知情的事件，导致输出格式受限，性能下降。为缓解此问题，我们提出不进行事件隐藏，而是生成自由格式的解释，但面临新挑战：大模型需抑制对角色未知事件的响应，因未隐藏信息使其全程可见。我们假设并实证发现，若在推理中显式揭示角色对事件的无知状态，大模型可实现有效抑制。基于此，我们提出PICTURE新提示方法，在自由格式思维链中显式表达角色的知识缺失。实验表明，PICTURE在虚假信念任务上平均比现有方法提升7.3%。
原文摘要 · Abstract (English)
Simulating human-like Theory of Mind (ToM) has been a longstanding problem in natural language processing (NLP). To address this, existing works introduce a reasoning step of event hiding (a.k.a. perspective-taking), where events unknown to a character are removed before question answering. However, resorting to event hiding for ToM reasoning presents a performance degradation issue due to the strict output format constraints involved in event hiding. To mitigate this issue, we propose generating perspective-taking outputs as free-form explanations without event hiding, but this poses a notable yet underexplored challenge: LLMs need to inhibit responses to events unknown to characters, because the absence of event hiding exposes LLMs to these events throughout reasoning. To address this challenge, we hypothesize and empirically verify that LLMs can achieve such inhibition if a character's lack of knowledge about events is made explicit during reasoning. Based on this finding, we introduce PICTURE, a new prompting method that enables LLMs to generate a character's lack of knowledge within free-form Chain-of-Thought (CoT). Experimental results show that PICTURE outperforms existing prompting methods by an average of 7.3% on false-belief tasks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。