Our latest work, Efficient Reinforcement Learning by Discovering Neural Pathways was accepted at NeurIPS 2024.