MemoryLACE: Memory Lifecycle-Aware Consolidation and Evidence Retrieval

2026-09-02

AI Summary

为解决长期语言模型记忆管理问题，提出MemoryLACE框架，通过稀疏合并、替代和矛盾关系显式建模文本证据生命周期，提高推理性能并减少运行时间。

Abstract

Long-term LLM agents must preserve information across interactions while distinguishing repeated evidence, historical states, updates, and unresolved contradictions. Existing textual memory systems retrieve semantically relevant memories efficiently but often leave these relationships implicit, whereas richer structured approaches model them through global graphs, hierarchical abstractions, or reflection at greater complexity. We introduce MemoryLACE (MemLACE), a lightweight memory framework that explicitly models the lifecycle of textual evidence through sparse merge, supersession, and contradiction relations while preserving atomic natural-language memories and their provenance. Rather than retrieving memories independently, MemLACE reconstructs relation-aware evidence units that expose current, historical, supporting, and conflicting evidence for downstream reasoning. Across BEAM and StructMemEval, using open-weight and proprietary LLM backbones, MemLACE achieves the highest overall performance in same-backbone comparisons while reducing end-to-end runtime on BEAM by 66.6% relative to Hindsight, the strongest reported reflective-memory baseline. Ablation studies identify lifecycle expansion and temporal awareness as the principal contributors to these gains. Together, the results demonstrate that explicitly modeling the local lifecycle of textual evidence is sufficient to substantially improve long-term memory reasoning without requiring comprehensive knowledge graphs or global reflection.

Problem

• long-term LLM agents
• textual memory systems
• semantic relevance
• structured approaches
• global graphs

Innovation

• sparse merge
• supersession
• contradiction relations
• relation-aware evidence units
• lifecycle expansion