Return to Article Details Beyond Hard Constraints: Budget-Conditioned Reachability for Safe Offline Reinforcement Learning Download Download PDF