Fetching the paper…

M3HF: Multi-agent Reinforcement Learning from Multi-phase Human Feedback of Mixed Quality · Around