FAITH: Feasibility-Aware Safety-Filtered RL for High-Dimensional Systems
Advances safe reinforcement learning for robotics with novel safety-filtering approach that handles infeasible states.
AI Summary
Researchers propose FAITH, a feasibility-aware safety-filtered RL framework that achieves high safety rates while maintaining performance on high-dimensional robotic systems.
Excerpt
Safe reinforcement learning commonly places safety and task performance in the same policy objective, where they can introduce competing updates. Safety filters separate them at action execution, but classical designs require an analytic safety function and dynamics model, and standard minimal-intervention filters are myopic to long-horizon task return because they minimize only instantaneous action deviation. Hard projections are also undefined when no safe action exists. We present FAITH, a fe