AI RESEARCH
Latent Q-Barrier Shielding for Safe In-Context Reinforcement Learning
arXiv CS.AI
•
ArXi:2605.25267v1 Announce Type: cross Safe in-context reinforcement learning (ICRL) adapts online from interaction history without test-time parameter updates while controlling episode cost under a safety budget. Under out-of-distribution (OOD) deployment shifts, pre