Safe Reinforcement Learning

Learning to Navigate Under Imperfect Perception: Conformalised Segmentation for Safe Reinforcement Learning

Safe Reinforcement Learning in Black-Box Environments via Adaptive Shielding