ResearchBenchmarksReinforcement Learning on Safety Gym AntVelFollow1RewardCOptiDICE-1.0904-0.5477-0.0050.5377Feb 19, 2026Evaluation ResultsMethodMethodLinksRewardCostCOptiDICE2026.0213.28BC-Safe2026.020.980.29LSPC-O2026.020.980.45LexiSafe-SC2026.020.980.73FISOR2026.020.890CPQ2026.02-1.010