2 papers
cs.LG2026
Safe Langevin Soft Actor Critic
Mahesh Keswani, Samyak Jain, Raunak P. Bhattacharyya
Balancing reward and safety in constrained reinforcement learning remains challenging due to poor generalization from sharp value minima and inadequate handling of heavy-tailed ris…
cs.LG2025
Learning Safe Autonomous Driving Policies Using Predictive Safety Representations
Mahesh Keswani, Raunak Bhattacharyya
Safe reinforcement learning (SafeRL) is a prominent paradigm for autonomous driving, where agents are required to optimize performance under strict safety requirements. This dual o…