LLMs能否通过反向表现模拟人物角色？一种针对数学推理情境中反事实指令跟随的系统性调查
Can LLMs Simulate Personas with Reversed Performance? A Systematic Investigation for Counterfactual Instruction Following in Math Reasoning Context
AI总结：
本文提出首个评估LLM反事实指令跟随能力的基准数据集，发现主流模型在模拟反向表现人物时表现不佳，揭示了反事实指令跟随的挑战。
摘要：
大型语言模型（LLMs）如今越来越多地用于虚拟环境中模拟人物角色，利用其指令跟随能力。然而，我们发现即使最先进的LLMs也无法模拟反向表现（如教育环境中低熟练度的学生角色），这影响了模拟的多样性并限制了模拟环境的实用性。在本工作中，以数学推理为代表性场景，我们提出了首个评估LLM在模拟反向表现人物能力的基准数据集，该能力我们称为“反事实指令跟随”。我们评估了开源和闭源LLMs在此任务上的表现，发现包括OpenAI o1推理模型在内的LLMs均难以遵循反事实指令来模拟反向表现的人物。同时模拟人物的表现水平和种族人口进一步恶化了效果。这些结果突显了反事实指令跟随的挑战，并指明了进一步研究的必要性。
英文摘要：
Large Language Models (LLMs) are now increasingly widely used to simulate personas in virtual environments, leveraging their instruction-following capability. However, we discovered that even state-of-the-art LLMs cannot simulate personas with reversed performance (e.g., student personas with low proficiency in educational settings), which impairs the simulation diversity and limits the practical applications of the simulated environments. In this work, using mathematical reasoning as a representative scenario, we propose the first benchmark dataset for evaluating LLMs on simulating personas with reversed performance, a capability that we dub "counterfactual instruction following". We evaluate both open-weight and closed-source LLMs on this task and find that LLMs, including the OpenAI o1 reasoning model, all struggle to follow counterfactual instructions for simulating reversedly performing personas. Intersectionally simulating both the performance level and the race population of a persona worsens the effect even further. These results highlight the challenges of counterfactual instruction following and the need for further research.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
