Approximate Model-Based Shielding for Safe Reinforcement Learning

Abstract

Reinforcement learning (RL) has shown great potential for solving complex tasks in a variety of domains. However, applying RL to safety-critical systems in the real-world is not easy as many algorithms are sample-inefficient and maximising the standard RL objective comes with no guarantees on worst-case performance. In this paper we propose approximate model-based shielding (AMBS), a principled look-ahead shielding algorithm for verifying the performance of learned RL policies w.r.t. a set of given safety constraints. Our algorithm differs from other shielding approaches in that it does not require prior knowledge of the safety-relevant dynamics of the system. We provide a strong theoretical justification for AMBS and demonstrate superior performance to other safety-aware approaches on a set of Atari games with state-dependent safety-labels.

arXiv commentAccepted at ECAI 2023 (main technical track)

References in corpus (26)

• Addressing Function Approximation Error in Actor-Critic Methods
• DeepMind Control Suite
• Recurrent World Models Facilitate Policy Evolution
• Learning Latent Dynamics for Planning from Pixels
• Implicit Quantile Networks for Distributional Reinforcement Learning
• Dopamine: A Research Framework for Deep Reinforcement Learning
• Benchmarking Batch Deep Reinforcement Learning Algorithms
• Lyapunov-based Safe Policy Optimization for Continuous Control
• Dream to Control: Learning Behaviors by Latent Imagination
• When to Trust Your Model: Model-Based Policy Optimization
• Mastering Diverse Domains through World Models
• Projection-Based Constrained Policy Optimization
• A Game Theoretic Framework for Model Based Reinforcement Learning
• The MineRL 2019 Competition on Sample Efficient Reinforcement Learning using Human Priors
• Shield Synthesis: Runtime Enforcement for Reactive Systems
• Mastering Atari with Discrete World Models
• IPO: Interior-point Policy Optimization under Constraints
• Constrained Model-based Reinforcement Learning with Robust Cross-Entropy Method
• Learning Belief Representations for Imitation Learning in POMDPs
• Constrained Policy Optimization via Bayesian World Models
• Safe Reinforcement Learning by Imagining the Near Future
• Learning Barrier Certificates: Towards Safe Reinforcement Learning with Zero Training-time Violations
• Do Androids Dream of Electric Fences? Safety-Aware Reinforcement Learning with Latent Shielding
• Learning in POMDPs is Sample-Efficient with Hindsight Observability
• Model-based Dynamic Shielding for Safe and Efficient Multi-Agent Reinforcement Learning
• Approximate Shielding of Atari Agents for Safe Exploration