为何会痛苦：识别情感支持对话中消极想法的驱动因素
Why It Hurts: Identifying the Drivers of Negative Thoughts in Emotional Support Conversations
- King’s College London（伦敦国王学院）
- JinFlow Intelligence Ltd.（金流智能公司）
- The Alan Turing Institute（阿兰·图灵研究所）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本研究针对LLM情感支持任务中认知评估维度显著性被忽视的问题，推出AppraiSal基准并提出基于贝叶斯逆规划的多智能体概率框架PRISM，提升了LLMs识别情境特定显著评估维度的能力。
AI中文摘要：
大型语言模型（LLMs）越来越多地被用于情感支持任务，例如消极想法重构。该任务依赖于修改认知评估，即对引发消极情绪的事件的主观解释，通常沿多个离散维度进行概念化。当前基于LLM的框架通过详尽评估所有可能维度来建模认知评估，但未考虑这些维度在不同情境下的不同显著性。在本研究中，我们探讨一个重要但被忽视的问题：“LLMs能否从情感支持对话中推断出显著的评估维度？”为解决该问题，我们推出AppraiSal基准，包含996个带有人类标注心理状态（包括显著认知评估维度）的情感支持对话。此外，我们提出PRISM，一种基于贝叶斯逆规划的多智能体概率框架，旨在提升LLMs识别情境特定评估维度的能力。实验结果显示，PRISM可提升不同规模LLMs的性能，尤其在识别最显著评估维度方面表现突出。
英文摘要：
Large Language Models (LLMs) are increasingly used for emotional support tasks, such as negative thought reframing. This task relies on modifying cognitive appraisals, the subjective interpretation of events that elicit negative emotions, which is typically conceptualized along multiple discrete dimensions. Current LLM-based frameworks model cognitive appraisal by exhaustively evaluating all possible dimensions, but they fail to account for the varying saliency of these dimensions across different contexts. In this work, we investigate a vital yet overlooked question: "Can LLMs infer the salient appraisal dimensions from emotional support conversations?" To address this question, we introduce the AppraiSal benchmark, containing 996 emotional support conversations with human-annotated mental states, including salient cognitive appraisal dimensions. Furthermore, we propose PRISM, a multi-agent probabilistic framework grounded in Bayesian Inverse Planning, designed to improve LLMs' ability to identify context-specific appraisal dimensions. Experimental results show that PRISM brings improvements to LLMs across various sizes, particularly in identifying the most salient appraisal dimensions.
