MemForest：一種採用分層時間索引的高效智能體記憶系統

摘要

記憶是實現長上下文大語言模型代理的基礎元件，透過持續的服務與更新生命週期，在跨互動過程中維持持久狀態。儘管已有大量前期研究，現有系統因兩項主要限制而面臨顯著的維護負擔：粗粒度的狀態管理與天生順序化的更新流程。具體而言，更新經常與大語言模型推理緊密耦合，且需要進行全狀態重寫，導致隨著記憶累積而產生較差的可擴展性與日益增加的延遲。為應對這些挑戰，我們提出 MemForest，一個將代理記憶重新構築為寫入高效的時序資料管理問題的記憶框架。MemForest 透過並行區塊提取打破順序瓶頸，將記憶建構解耦為可並行運行的獨立操作。為進一步消除粗粒度維護，我們引入 MemTree，一種層級時間索引，將記憶組織為按時間排序的樹狀結構，而非平坦的全局摘要。此設計以局部化的逐節點更新取代全狀態重寫，將維護成本降低至受影響的樹路徑，同時自然保留隨時間演變的狀態。我們在兩個長上下文記憶基準 LongMemEval-S 與 LoCoMo 上評估 MemForest。在 LongMemEval-S 中，MemForest 在有狀態基線中達到最佳整體表現，實現 79.8% 的 pass@1 準確率，同時記憶建構吞吐量比包括 EverMemOS 在內的最新方法高出約 6 倍。

English

Memory is a fundamental component for enabling long-context LLM agents, supporting persistent state across interactions through a continuous serve-and-update lifecycle. Despite substantial prior work, existing systems suffer from significant maintenance overhead due to two key limitations: coarse-grained state management and inherently sequential update pipelines. In particular, updates are often tightly coupled with LLM inference and require full-state rewrites, leading to poor scalability and growing latency as memory accumulates. To address these challenges, we present MemForest, a memory framework that reformulates agent memory as a write-efficient temporal data management problem. MemForest breaks the sequential bottleneck via parallel chunk extraction, decoupling memory construction into concurrent, independent operations. To further eliminate coarse-grained maintenance, we introduce MemTree, a hierarchical temporal index that organizes memory as time-ordered trees rather than flat global summaries. This design replaces full-state rewrites with localized per-node updates, reducing maintenance cost to the affected tree paths while naturally preserving temporally evolving states. We evaluate MemForest on two long-context memory benchmarks, LongMemEval-S and LoCoMo. On LongMemEval-S, MemForest achieves the best overall performance among stateful baselines, reaching 79.8% pass@1 accuracy while sustaining a memory construction throughput approximately 6x higher than state-of-the-art approaches including EverMemOS.