AI 的记忆正在害它自己，这篇给它装了回查档案
AI 智能体靠「记忆」变聪明，但记忆也会让它变蠢：明明没有记忆时能解开的题，一翻旧账反而做错——因为它把一条错误经验当真理，或者把对的用错了地方。问题出在经验是从任务结果反推的，却没附上原始过程，AI 无从核对。这篇 WikiTrace 的做法很朴素：每条经验都挂回它来源的完整轨迹和成败原因，AI 动手前先查档，发现经验缺细节、互相矛盾、或来自失败任务时，就顺着链接翻原始记录修正。在 5 个智能体基准上，它比现有记忆方案和「只给经验不给档案」的对照组都稳。它不是你明天能用上的东西，但它点破了一个正在发生的隐患：我们越依赖 AI 的长期记忆，越要给它留一条回到事实的路。
📄 原文摘要(英文)
External memory enables LLM agents to improve from experience by distilling trajectories of past tasks into concise lessons for new tasks. However, memory can also lead to memory-induced degradation, where an agent fails on tasks it would solve without memory, often because it follows a lesson that is incorrect, or misapplies or misreads a valid one. This is because lessons are inferred from trajectories labeled only with task success or failure, yet are typically provided without the trajectories they came from, leaving the agent no way to check them. To address this, we introduce WikiTrace, an evidence-linked memory system designed to let agents check each lesson against its original trajectory. Specifically, motivated by the finding that a trajectory often still holds the evidence needed to catch an erroneous lesson, WikiTrace stores each lesson as a concise insight linked to a summary of why its trajectory succeeded or failed and to the raw trajectory itself. Then, on a new task, the agent starts from all insights and, whenever one lacks details needed to act, conflicts with another, or comes from a failed task, descends through its links with terminal tools to fill in missing details or correct the insight before acting. We validate WikiTrace on five agentic benchmarks, showing that it consistently outperforms existing memory approaches and a baseline given the same insights alone, while also reducing memory-induced degradation. Our code is available at https://github.com/yukyeongleee/wikitrace