给 AI 装记忆，越复杂越好？这篇说：未必
给 AI 装记忆，大家默认越复杂越好：要整理、要分层、要学怎么检索。这篇把最笨的办法——把聊天记录原封不动存着，要用时直接翻——拿去和一堆花哨的「记忆系统」比，结果在 7 项测试里赢了 6 项。原因也简单：花哨系统整理记忆时，把关键信息弄丢了；原样保存的，反而找得全。它不是你明天能用上的，但提醒了一件事：在给 AI 加复杂度之前，先确认它是不是真的缺那层复杂度。
📄 原文摘要(英文)
LLM memory is widely used in long-term conversation understanding. Agentic memory systems often use mutation, structural organization, learned retrieval strategies, or other complex operations. Is this complexity always necessary? We study Raw-RAG, a naive RAG baseline using dense retrieval and reranking over raw history to supply evidence to a reader LLM. Across LoCoMo, LongMemEval, MEME, and MemoryAgentBench, Raw-RAG ranks first among the compared memory systems in six of seven benchmark-model settings, outperforming the strongest compared agentic memory system by 0.13-5.53 points. We investigate why raw-history retrieval remains competitive. Evidence analysis shows that its higher coverage of necessary evidence can offset lower utilization. Useful evidence includes both answer values and the relations and dependency paths needed to derive them. Even complete access to transformed memory stores leaves an accuracy gap relative to raw evidence, highlighting the importance of preserving raw information. These findings challenge the necessity of complex agentic memory for the evaluated conversation tasks and establish naive RAG as an essential baseline.