analiza aiqbench
MACE introduces a novel framework for multi-agent systems that leverages execution feedback to dynamically organize memory into functional units via adaptive graphs. By optimizing how agents retrieve and utilize past collaboration traces, this approach achieves significant performance gains across diverse benchmarks compared to existing baselines. This advancement directly addresses the critical challenge of scaling LLM-based agent teams through more efficient knowledge retention and retrieval mechanisms.
MACE: Memory-Agent Co-Evolution with Adaptive Memory Graphs for Multi-Agent Systems
arXiv:2609.21533v1 Announce Type: new Abstract: LLM-based multi-agent systems generate collaboration traces that record how agents plan tasks, verify intermediate results, and repair failures. Reusing these procedures requires preserving an action's prerequisites and the outputs needed by subsequent agents. Our empirical studies show that grouping these dependencies into functional memory units improves their retention, while connecting units increases retrieval of the units and links jointly required by a task. The preferred combination of units also changes between instructions and checklists, even when each combination's content is fixed across formats. Updating choices from the outcomes of each combination and format pairing outperforms scoring combinations and formats separately. These findings motivate MACE, a memory-agent co-evolution framework that adapts memory organization and agent memory use through execution feedback. Its MemGoG structure represents functional units as subgraphs of related conditions, actions, and outputs, connecting them through support, conflict, and repair relations. MACE Loop selects task-relevant units and relations within a memory budget and provides each agent with instructions or checklists for its current operation. It records the selected units, presentation formats, agent outputs, and task outcomes to update unit scores and relations for retrieval and inform subsequent presentation choices. Across eight benchmarks, MACE outperforms ten baselines with an average score of 81.11%, compared with 78.97% for the strongest baseline, SAGE.
Abstract:LLM-based multi-agent systems generate collaboration traces that record how agents plan tasks, verify intermediate results, and repair failures. Reusing these procedures requires preserving an action's prerequisites and the outputs needed by subsequent agents. Our empirical studies show that grouping these dependencies into functional memory units improves their retention, while connecting units increases retrieval of the units and links jointly required by a task. The preferred combination of units also changes between instructions and checklists, even when each combination's content is fixed across formats. Updating choices from the outcomes of each combination and format pairing outperforms scoring combinations and formats separately. These findings motivate MACE, a memory-agent co-evolution framework that adapts memory organization and agent memory use through execution feedback. Its MemGoG structure represents functional units as subgraphs of related conditions, actions, and outputs, connecting them through support, conflict, and repair relations. MACE Loop selects task-relevant units and relations within a memory budget and provides each agent with instructions or checklists for its current operation. It records the selected units, presentation formats, agent outputs, and task outcomes to update unit scores and relations for retrieval and inform subsequent presentation choices. Across eight benchmarks, MACE outperforms ten baselines with an average score of 81.11%, compared with 78.97% for the strongest baseline, SAGE.
Machine Learning (cs.LG)
arXiv:2609.21533 [cs.LG]
(or arXiv:2609.21533v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.21533 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 304 cuvinte
Subiecte conexe
Graph-Based Personalized Memory for LLM Agents: Representation, Evolution, Retrieval, and Evaluation
arXiv · 9 sept. 2026
Self-Evolving Agents as Dynamic Graph Transformation: A Survey and New Perspective
arXiv · 21 aug. 2026
EMIR$^2$: Evolution-Aware Memory with Intent-Guided Multi-Round Retrieval
arXiv · 29 sept. 2026
AI and Quantum Computing Briefing
livetradingnews.com · 4 aug. 2026