AI 中文总结
EngramRAG提出基于CLS的动态记忆架构，通过U-PPR、CATD和SUPERSEDES解决多跳智能体记忆缺陷，在LoCoMo上显著提升检索性能并消除幻觉。
AI 中文摘要
随着自主LLM智能体在多会话环境中部署，传统记忆架构面临联想盲视（无法遍历多跳关系依赖）、脚手架遗忘症（时间衰减驱逐核心人格不变量）和静态拓扑停滞（不可变图忽略使用动态）等问题。基于互补学习系统（CLS）原理，我们提出EngramRAG，一种自适应记忆架构，将低延迟的清醒状态反射与异步后台的梦境状态巩固周期相结合。EngramRAG引入：（1）使用调制个性化PageRank（U-PPR），其中转移概率通过赫布可塑性适应，将持久实体提升为高中心性的认知宏观枢纽；（2）巩固激活拓扑衰减（CATD），其将保留半衰期按拓扑承重权重而非挂钟新鲜度缩放，并由冷启动宽限期（N_grace >= 4）保护；（3）有向SUPERSEDES DAG过滤，以在事实突变期间抑制过时状态；（4）三源混合检索，通过动态倒数排名融合（RRF）融合稠密向量、BM25和U-PPR。在LoCoMo基准的10个长期对话中全部1,982个问答对上评估，EngramRAG在Recall@5上实现+38.9%的相对改进（53.21%对38.29%，p < 0.001），在MRR上实现+43.1%的改进（0.4203对0.2937），优于稠密向量RAG，显著优于Okapi BM25（48.66%）和孤立静态图检索（8.50%）。在时间推理上，EngramRAG达到62.33%的Recall@5（比稠密向量高+16.67个百分点）。在受控突变测试中，SUPERSEDES将分裂脑幻觉从70.0%抑制至0.0%，而90天模拟显示在26.21ms交互检索反射下实现100.0%的脚手架保留。
英文摘要
As autonomous LLM agents are deployed across multi-session environments, conventional memory architectures suffer from Associative Blindness (inability to traverse multi-hop relational dependencies), Scaffolding Amnesia (temporal decay evicting core persona invariants), and Static Topology Stagnation (immutable graphs ignoring usage dynamics). Grounded in Complementary Learning Systems (CLS) principles, we propose EngramRAG, an adaptive memory architecture coupling a low-latency Waking State reflex with an asynchronous background Dreaming State consolidation cycle. EngramRAG introduces: (1) Usage-Modulated Personalized PageRank (U-PPR), where transition probabilities adapt via Hebbian plasticity to promote persistent entities into high-centrality Epistemic Macro-Hubs; (2) Consolidation-Activated Topology Decay (CATD), which scales retention half-life by topological load-bearing weight rather than wall-clock recency, protected by a cold-start grace period (N_grace >= 4); (3) Directed SUPERSEDES DAG filtering to suppress obsolete state during fact mutations; and (4) Triple-source hybrid retrieval fusing dense vectors, BM25, and U-PPR via dynamic Reciprocal Rank Fusion (RRF). Evaluating on all 1,982 QA pairs across 10 long-term conversations in the LoCoMo benchmark, EngramRAG achieves +38.9% relative improvement in Recall@5 (53.21% vs. 38.29%, p < 0.001) and +43.1% in MRR (0.4203 vs. 0.2937) over dense vector RAG, significantly outperforming Okapi BM25 (48.66%) and isolated static graph retrieval (8.50%). On temporal reasoning, EngramRAG reaches 62.33% Recall@5 (+16.67 points over dense vectors). In controlled mutation tests, SUPERSEDES suppresses split-brain hallucinations from 70.0% to 0.0%, while 90-day simulations show 100.0% scaffolding retention under a 26.21ms interactive retrieval reflex.
Comments8 pages, 6 figures, 4 tables. Code and reproduction suite: https://github.com/bpoti001/epigraph
