CoMem: Collective-Individual Memory Synergy for Evolutionary Multi-Agent Systems
cs.AI
Submitted: 2026-09-14
Updated: 2026-09-14
Comments: Submitted to AAAI 2027.9 pages,4 figures
License: http://creativecommons.org/licenses/by/4.0/
The gist: Designing effective memory mechanisms is crucial for advancing LLM-driven Multi-Agent Systems (MAS), helping agents learn together and perform better over time.
Terminology
Abstract
Designing effective memory mechanisms is crucial for advancing LLM-driven Multi-Agent Systems (MAS), helping agents learn together and perform better over time. While recent work has led to strong cooperation skills, most methods still use flat, unstructured memories, which easily get filled with noise and erase differences between agents. To address this, we introduce the concept of collective-individual memory synergy and propose CoMem, an architecture that unifies both private experience and shared knowledge for multi-agent learning. CoMem features:(i) Private Experience Sedimentation, which lets each agent keep and update its own useful memories over time;(ii) Collective Wisdom Curation, which carefully selects only widely proven ideas to be shared among agents;(iii)Parallel Dual-Stream Retrieval, which allows agents to draw both from their own memory and the group's wisdom, using clustering to ensure diversity.Experiments on ALFWorld and PDDL benchmarks show that CoMem achieves strong overall performance and robustly avoids memory pollution.
Sources
- Collective memory, consensus, and learning explained by social cohesion
- Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory
- DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
- Memory for Autonomous LLM Agents:Mechanisms, Evaluation, and Emerging Frontiers
- LLM Multi-Agent Systems: Challenges and Open Problems
- Emergent Collective Memory in Decentralized Multi-Agent AI Systems
- RCR-Router: Efficient Role-Aware Context Routing for Multi-Agent LLM Systems with Structured Memory
- Choosing How to Remember: Adaptive Memory Structures for LLM Agents
- MemGPT: Towards LLMs as Operating Systems
- Collaborative Memory: Multi-User Memory Sharing in LLM Agents with Dynamic Access Control
- Memory poisoning and secure multi-agent systems
- Voyager: An Open-Ended Embodied Agent with Large Language Models
- Mixture-of-Agents Enhances Large Language Model Capabilities
- State Contamination in Memory-Augmented LLM Agents
- Qwen3 Technical Report
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
- VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection