SMOSE: Sparse Mixture of Shallow Experts for Interpretable Reinforcement Learning in Continuous Control Tasks Paper • 2412.13053 • Published Dec 17, 2024