让大模型像人一样基于情绪和未解冲突选择记忆。
PsychoAgent: An Affect-Sensitive Cognitive Architecture for Conflict-Aware Memory in LLM Agents
- 分离事实与情感记忆，用冲突感知控制器整合
- 在三种冲突场景中检索关键记忆准确率达0.933
- 适合研究人机认知、情感计算的开发者
人类认知不仅依据主题相似性选择经验：情感重要性和未解决冲突也影响记忆可及性。我们提出PsychoAgent，一种面向大模型智能体的认知架构，将事实记忆与情感记忆分离，并通过冲突感知执行控制器融合二者。情感记忆先经语义相关性过滤，再按显著性重排序，在保持主题匹配的同时，允许情感重要的记忆进入提示。在三个受控冲突场景中，完整架构检索到的冲突关键记忆占比达0.933，显著高于语义-情感和单记忆RAG基线（0.500和0.667），仅略有语义相似性损失。五名盲评者评估27个输出，经组内标准化后，完整架构总体均分最高（+0.22标准差），但成对差异不显著。三天示例轨迹显示持续情感影响、离线记忆重组与选择性重加权。结果支持情感敏感检索作为可解释机制，用于建模大模型智能体中的类人冲突效应。
原文摘要 · Abstract (English)
Human-like cognition does not select past experience by topical similarity alone: affective significance and unresolved conflict also shape what becomes accessible. We present PsychoAgent, a cognitive architecture for LLM agents that separates factual and affective memory and integrates both through a conflict-aware executive controller. Affective memories are first filtered by semantic relevance and then re-ranked by salience, preserving topical fit while allowing emotionally important traces to enter the prompt. Across three controlled conflict scenarios, the full architecture retrieved more conflict-critical memories than semantic-affective and single-memory RAG baselines (0.933 vs. 0.500 and 0.667), with a small semantic-similarity cost. Five blinded raters evaluated 27 outputs. After within-rater standardization, the full architecture had the highest overall mean (+0.22 SD), but corrected pairwise differences were not significant. A three-day illustrative trace further shows persistent affect, offline memory recombination, and selective memory reweighting. The findings support affect-sensitive retrieval as an inspectable mechanism for modeling human-like conflict effects in LLM agents.
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。