MemCollab: Cross-Agent Memory Collaboration via Contrastive Trajectory Distillation

2026-03-24

AI Summary

This work addresses the limitations of existing memory mechanisms in large language model agents, which are typically designed for single-agent settings and struggle to support effective knowledge sharing across heterogeneous agents, often leading to performance degradation when directly transferred. To overcome this, the authors propose a cross-agent collaborative memory framework that extracts task-level invariant reasoning constraints by contrasting the reasoning trajectories of diverse agents on identical tasks. Through contrastive trajectory distillation, the framework constructs a debiased, agent-agnostic shared memory representation. Coupled with a task-aware memory retrieval mechanism, this approach enhances both reasoning efficiency and accuracy. Experimental results demonstrate significant performance improvements across multiple agent types on mathematical reasoning and code generation tasks, validating the effectiveness of shared memory in cross-model-family scenarios.

Technology Category

• Multiagent Systems: Modeling other Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Large language model (LLM)-based agents rely on memory mechanisms to reuse knowledge from past problem-solving experiences. Existing approaches typically construct memory in a per-agent manner, tightly coupling stored knowledge to a single model's reasoning style. In modern deployments with heterogeneous agents, a natural question arises: can a single memory system be shared across different models? We found that naively transferring memory between agents often degrades performance, as such memory entangles task-relevant knowledge with agent-specific biases. To address this challenge, we propose MemCollab, a collaborative memory framework that constructs agent-agnostic memory by contrasting reasoning trajectories generated by different agents on the same task. This contrastive process distills abstract reasoning constraints that capture shared task-level invariants while suppressing agent-specific artifacts. We further introduce a task-aware retrieval mechanism that conditions memory access on task category, ensuring that only relevant constraints are used at inference time. Experiments on mathematical reasoning and code generation benchmarks demonstrate that MemCollab consistently improves both accuracy and inference-time efficiency across diverse agents, including cross-modal-family settings. Our results show that the collaboratively constructed memory can function as a shared reasoning resource for diverse LLM-based agents.

Problem

• memory sharing
• heterogeneous agents
• agent-specific bias
• cross-agent collaboration
• reasoning invariants

Innovation

• cross-agent memory
• contrastive trajectory distillation
• agent-agnostic memory
• reasoning constraints
• task-aware retrieval