TY - RPRT TI - Tackling the Zero-Shot Reinforcement Learning Loss Directly AU - Yann Ollivier PY - 2025 UR - https://arxiv.org/abs/2502.10792 ID - 2502.10792 ER -