发现记忆保留前就失败的隐性问题，提出解决方法提升大模型记忆准确率。
When Retrieval Fails Before It Begins: Structurally Indirect Prerequisite Eviction as a Retention Failure in Agentic Memory
- 识别出上游弱相关块被误删导致检索失败的新机制
- 新方法使记忆保留率从0.03提升至0.90（词编码器）
- 适用于需精准记忆的复杂推理任务
代理记忆在固定预算下分为保留与检索两个阶段。现有以检索为中心的范式隐含假设必要证据不会被丢弃，但我们揭示了一种预检索失败模式：结构间接前提剔除，即在预算压力下，与查询弱对齐的上游块被错误淘汰。本文提出该失败的操作定义、可复现的确定性基准及按种子追踪诊断方法。进一步评估了依赖感知语义垃圾回收（DSGC）——一种单跳图感知规则。在主实验中，DSGC将全链路记忆保留率从词编码器下的0.03提升至0.90，句子编码器下从0.23提升至1.00。稳健性检验还明确了该规则有效或退化的预算与扩展范围。所发布的流程与故障回溯分析支持对检索前保留失败这一独立失效边界进行机制研究。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。