TY - RPRT TI - Exploration is Harder than Prediction: Cryptographically Separating Reinforcement Learning from Supervised Learning AU - Noah Golowich AU - Ankur Moitra AU - Dhruv Rohatgi PY - 2024 UR - https://arxiv.org/abs/2404.03774 ID - 2404.03774 ER -