CoRe-MARL: Cooperative Redistribution Under Unknown Dynamics Using Recurrent Multi-Agent Reinforcement Learning
A multi-agent reinforcement learning method for cooperative redistribution under unknown dynamics, relevant to AI capability research.
A multi-agent reinforcement learning method for cooperative redistribution under unknown dynamics, relevant to AI capability research.
Auto-scored: A technical arXiv paper on multi-agent RL, a minor capability advance with no direct misuse or governance implications.