当检索尚未开始时检索就已失败：智能体记忆中作为保留失败的结构间接先决条件驱逐
When Retrieval Fails Before It Begins: Structurally Indirect Prerequisite Eviction as a Retention Failure in Agentic Memory
浏览论文内容
中文总结 AI 辅助
本文针对智能体记忆的检索前保留失败问题，提出结构间接先决条件驱逐的概念，评估DSGC规则显著提升全链保留率，并发布相关工具支持机制分析。
中文摘要 AI 辅助
固定预算下的智能体记忆包含保留与检索两个阶段。现有以检索为中心的范式默认必要证据会在驱逐中保留，但本文对此提出挑战，分离出一种检索前失败模式：结构间接先决条件驱逐，即与查询弱对齐的上游块在预算压力下被丢弃。本文给出该失败的可操作定义、可复现的确定性基准及逐种子轨迹诊断，最后评估了Dependency-aware Semantic Garbage Collection（DSGC）这一基于单跳图的规则。在主实验套件中，DSGC使全链保留率在词法编码器下从0.03提升至0.90，在句子编码器下从0.23提升至1.00；鲁棒性检查还确定了单跳规则成立或退化的预算与规模区间，所发布的流水线与失败事后分析为研究检索前的保留提供了新的失败边界。
英文摘要
Agentic memory under a fixed budget involves two stages: retention and retrieval. Existing retrieval-centered paradigms implicitly assume necessary evidence survives eviction, but we challenge this by isolating a pre-retrieval failure mode: structurally indirect prerequisite eviction, in which upstream blocks weakly aligned with the query are discarded under budget pressure. We provide an operational definition of this failure, a reproducible deterministic benchmark, and per-seed trace diagnostics. Finally, we evaluate Dependency-aware Semantic Garbage Collection (DSGC), a one-hop graph-aware rule. In our main suite, DSGC improves full-chain retention from 0.03 to 0.90 under a lexical encoder and from 0.23 to 1.00 under a sentence encoder. Robustness checks then identify the budget and scaling regimes where the one-hop rule holds or degrades. Our released pipeline and failure postmortem support mechanistic analysis of retention before retrieval as a distinct failure boundary.
