REVIEW 2 cited by
Safe Reinforcement Learning via Probabilistic Shields
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
This paper targets the efficient construction of a safety shield for decision making in scenarios that incorporate uncertainty. Markov decision processes (MDPs) are prominent models to capture such planning problems. Reinforcement learning (RL) is a machine learning technique to determine near-optimal policies in MDPs that may be unknown prior to exploring the model. However, during exploration, RL is prone to induce behavior that is undesirable or not allowed in safety- or mission-critical contexts. We introduce the concept of a probabilistic shield that enables decision-making to adhere to safety constraints with high probability. In a separation of concerns, we employ formal verification to efficiently compute the probabilities of critical decisions within a safety-relevant fragment of the MDP. We use these results to realize a shield that is applied to an RL algorithm which then optimizes the actual performance objective. We discuss tradeoffs between sufficient progress in exploration of the environment and ensuring safety. In our experiments, we demonstrate on the arcade game PAC-MAN and on a case study involving service robots that the learning efficiency increases as the learning needs orders of magnitude fewer episodes.
Forward citations
Cited by 2 Pith papers
-
Conformal Safety Shielding for Imperfect-Perception Agents
The paper introduces a conformal prediction-based shield for imperfect-perception agents and proves a global safety bound only for the simpler perfect-perception case.
-
Efficient Dynamic Shielding for Parametric Safety Specifications
Dynamic shields precompute atomic safety controllers offline and compose them online, so a changing safety specification can be enforced in seconds instead of recomputing a full shield.
Discussion (0). Continue with ORCID to comment.