FlexMARL: Graph-Driven Rollout-Training Co-Design for LLM-Based Multi-Agent Reinforcement Learning
Open Access
- 14 November 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1371-1386
Abstract
No abstract availableThis publication has 25 references indexed in Scilit:
- LERO: LLM-driven Evolutionary Framework with Hybrid Rewards and Enhanced Observation for Multi-Agent Reinforcement Learning2025
- MAGE: A Multi-Agent Engine for Automated RTL Code Generation2025
- LAMARL: LLM-Aided Multi-Agent Reinforcement Learning for Cooperative Policy GenerationIEEE Robotics and Automation Letters, 2025
- A survey on LLM-based multi-agent systems: workflow, infrastructure, and challengesVicinagearth, 2024
- Colossal-AI: A Unified Deep Learning System For Large-Scale Parallel Training2023
- Cooperative Learning of Multi-Agent Systems Via Reinforcement LearningIEEE Transactions on Signal and Information Processing over Networks, 2023
- Multi-Agent Reinforcement Learning: A Selective Overview of Theories and Algorithms2021
- ZeRO: Memory optimizations Toward Training Trillion Parameter Models2020
- MARL-Ped: A multi-agent reinforcement learning based framework to simulate pedestrian groupsSimulation Modelling Practice and Theory, 2014
- Improved utilization and responsiveness with gang scheduling1997