让记忆提前推理，让对话更省力
Pre-Storage Reasoning for Episodic Memory: Shifting Inference Burden to Memory for Personalized Dialogue
- 记忆构建时预先完成复杂推理，减轻生成负担
- 小模型性能接近大模型，且节省计算资源
- 适合需要长期记忆的个性化对话系统
对话型AI实现有效长时记忆，需融合多轮会话信息。但现有系统将大量推理压力置于响应生成阶段，导致性能严重依赖模型规模。我们提出PREMem（预存储推理机制），将复杂推理过程从推理阶段转移到记忆构建阶段。PREMem提取细粒度的记忆片段，分为事实、体验和主观三类，并建立跨会话的记忆项间显式关联，捕捉扩展、转化、推论等演化模式。通过在预存储阶段完成推理，PREMem生成丰富表征的同时，显著降低交互时的计算开销。实验表明，所有模型规模下性能均有显著提升，小模型表现可媲美更大规模基线，且在有限令牌预算下仍保持有效性。代码与数据集见https://github.com/sangyeop-kim/PREMem。
原文摘要 · Abstract (English)
Effective long-term memory in conversational AI requires synthesizing information across multiple sessions. However, current systems place excessive reasoning burden on response generation, making performance significantly dependent on model sizes. We introduce PREMem (Pre-storage Reasoning for Episodic Memory), a novel approach that shifts complex reasoning processes from inference to memory construction. PREMem extracts fine-grained memory fragments categorized into factual, experiential, and subjective information; it then establishes explicit relationships between memory items across sessions, capturing evolution patterns like extensions, transformations, and implications. By performing this reasoning during pre-storage rather than when generating a response, PREMem creates enriched representations while reducing computational demands during interactions. Experiments show significant performance improvements across all model sizes, with smaller models achieving results comparable to much larger baselines while maintaining effectiveness even with constrained token budgets. Code and dataset are available at https://github.com/sangyeop-kim/PREMem.
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。