Pieter Abbeel - Understanding Sparse Rewards in RL (via searchlore.ai)