长程语言智能体中的重尾记忆痕迹与CTWM控制器
原标题：Heavy-Tailed Memory Traces in Long-Horizon Language Agents
AI 摘要
Emory大学与东京大学的研究者提出对长程语言智能体外部记忆的「重尾痕迹」审计方法，发现有限上下文下记忆会集中于小核心并留下长尾，且该分布随策略不同而变化。基于此提出Core–Tail World Model（CTWM）控制器，按排名分配提示预算并保留摘要尾部。在Synthetic Graph World上提示token减少5.9%、尾部预测误差降低13.6%，在LongMemEval上token减少24.48%且准确率持平。
正文节选
Heavy-Tailed Memory Traces in Long-Horizon Language Agents Abstract Long-horizon language agents increasingly rely on external memory as a frozen world model, yet current memory systems are usually judged only by task success or token cost. We argue that the missing object is the shape of memory use: under finite context and repeated retrieval, agent memory can concentrate on a small core while leaving rare states in a long tail where prediction errors accumulate. We study this effect through a