Postdoc, Nanjing university
1 paper at NeurIPS 2025
A novel value decomposition framework of Continued Fraction Q-Learning (QCoFr) is proposed to model rich cooperation for multi-agent reinforcement learning without combinatorial explosion.
