TY - RPRT TI - From Novelty to Imitation: Self-Distilled Rewards for Offline Reinforcement Learning AU - Gaurav Chaudhary AU - Laxmidhar Behera PY - 2025 UR - https://arxiv.org/abs/2507.12815 ID - 2507.12815 ER -