Synthesizing Safe Policies under Probabilistic Constraints with Reinforcement Learning and Bayesian Model Checking
This addresses safety-critical domains for reinforcement learning practitioners, but it appears incremental as it builds on existing methods for constraint specification and Bayesian model checking.
The paper tackled the problem of ensuring safety in reinforcement learning under probabilistic constraints by leveraging epistemic uncertainty about constraint satisfaction, and the result was a framework that uses an agent's confidence in constraint satisfaction to balance optimization and safety.
We propose to leverage epistemic uncertainty about constraint satisfaction of a reinforcement learner in safety critical domains. We introduce a framework for specification of requirements for reinforcement learners in constrained settings, including confidence about results. We show that an agent's confidence in constraint satisfaction provides a useful signal for balancing optimization and safety in the learning process.