Reward hacking in physical reinforcement learning revealed by turbulent drag reduction

Excerpt

arXiv:2606.06227v3 Announce Type: replace-cross Abstract: Reinforcement-learning controllers optimise specified rewards, but in physical systems those rewards often capture only part of the true control objective. Three mechanisms through which this mismatch can produce apparent success without physical improvement are identified: incomplete accounting that omits relevant costs, constraint enforcement outside the policy that corrupts credit assignment, and observations that fail to resolve the r