发表机构
South Dakota State University（南达科他州立大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对LLM智能体的内存滞后问题，提出弹性内存组件CrystalMem，通过知识结晶策略恢复能力，在多环境实验中性能优于基准方法。
AI 中文摘要
自进化大语言模型（LLM）智能体的内存通常被配置为仅按字节预算增长。然而，云平台会根据负载和成本调整配额，我们发现其能力并不会随预算恢复而回升：在经历压缩-恢复循环后，智能体的能力会低于压缩前的水平，我们将这一差距称为内存滞后。其原因是结构性的：删除和单向压缩会丢弃后续重建所需的内容，且我们证明，任何仅保留或丢弃条目的策略都会存在残余缺陷下限。我们提出CrystalMem（结晶内存），这是一种弹性内存辅助组件，它根据结晶能量计划将条目降级为四个保真度状态，通过优势加权影响力结合依赖耦合对降级进行排序，并在明确的计算和字节限制下通过验证重结晶恢复能力。在七个环境、十七种方法和六个主干模型的实验中，结合多租户服务和物理边缘-云部署，CrystalMem在所有设置中均实现了最高的恢复能力，填补了所有基准方法留下的空白：从50%的字节预算开始，CrystalMem在每个环境中都匹配了全预算基准的最强能力；在预算相等的情况下，其平均领先4.6个百分点。
英文摘要
Memory for self-evolving large language model (LLM) agents is often provisioned as if its byte budget only grows. Cloud platforms, however, adjust quotas with load and cost, and we show that capability does not follow the budget back up: after a squeeze-and-recover cycle, the agent settles below its pre-squeeze level, a gap we call memory hysteresis. The cause is structural. Deletion and one-way compression discard the material needed for later rebuilding, and we prove that any policy that only keeps or drops entries carries a residual-deficit floor. We propose CrystalMem (Crystallized Memory), an elastic memory sidecar that demotes entries across four fidelity states under a crystallization-energy schedule, orders demotions by advantage-weighted influence with dependency coupling, and recovers capability through verified recrystallization under explicit compute and byte caps. Across seven environments, seventeen methods, and six backbones, with multi-tenant serving and a physical edge-cloud deployment, CrystalMem achieves the highest restored capability in every setting and closes the loop left open by every baseline. From a 50% byte budget, CrystalMem matches the strongest budgeted baseline at full provision on every environment; at equal budgets, it leads by +4.6 pp on average.
