PhD student, Institute of automation, Chinese academy of science
1 paper at NeurIPS 2025
We propose a novel framework employs a MoE-based multi-agent world model for learning and planning in multi-task scenarios.
