Agents
MemForest: Efficient Agent Memory Management via EventTree Partitioning and Progressive Merging
arXiv:2609.08273v1 Announce Type: new Abstract: Agent memory systems have demonstrated significant potential in long-term dialogue, personalized assistants, and video understanding. However, continuou
arXiv:2609.08273v1 Announce Type: new Abstract: Agent memory systems have demonstrated significant potential in long-term dialogue, personalized assistants, and video understanding. However, continuously accumulated memory introduces substantial storage and retrieval costs during inference. To address this issue, we propose extbf{MemForest}, a general memory compression framework adaptable to various agent memory systems. Specifically, MemForest partitions historical memory into event-centric units by leveraging global semantic similarity and local temporal continuity. For each unit, it constructs a maximum spanning tree, termed an EventTree, and progressively merges redundant memory nodes by selecting high-weight edges, reducing storage overhead. Furthermore, we introduce an anchor-guided propagation retrieval mechanism that retrieves relevant memory nodes from the temporal neighborhoods of key nodes, improving retrieval accuracy. Extensive experiments demonstrate the effectiveness of MemForest. Under the unimodal Mem0 framework, MemForest retains extbf{97.1%} of the original performance while compressing extbf{50%} of historical memory across three benchmarks (LoCoMo, LongMemEval, and PersonaMem), achieving a extbf{1.89x} retrieval speedup. Under the multimodal M3-Agent framework, it preserves extbf{99.7%} of the original performance with a extbf{50%} compression ratio across two benchmarks (M3-Bench-robot and M3-Bench-web), achieving a extbf{2.24x} retrieval speedup. extcolor{RoyalBlue}{extit{Our code is available at https://github.com/Celina-love-sweet/MemForest.}}
Related
- CMI-Mem: Toward Generalizable Long-Term Memory Management via CMI-Augmented Reinforcement Learning
- MemForest: An Efficient Agent Memory System with Hierarchical Temporal Indexing
- Infini Memory: Maintainable Topic Documents for Long-Term LLM Agent Memory
- GraphMemix: Query-Aware Evidence Forests for Long-Term Multimodal Agent Memory
- Mandol: An Agglomerative Agent Memory System for Long-Term Conversations
Source: arXiv cs.AI | 2026-09-10