mlsanityData to 7 Oct 2026

Provable Safe Reinforcement Learning with Binary Feedback

arXiv 2210.14492 ↗First posted 26 Oct 2022cs.LG cs.AI stat.MLcited by 7

Not on the map. Why?

Abstract (arXiv)

Safety is a crucial necessity in many applications of reinforcement learning (RL), whether robotic, automotive, or medical. Many existing approaches to safe RL rely on receiving numeric safety feedback, but in many cases this feedback can only take binary values; that is, whether an action in a given state is safe or unsafe. This is particularly true when feedback comes from human experts. We therefore consider the problem of provable safe RL when given access to an offline oracle providing binary feedback on the safety of state, action pairs. We provide a novel meta algorithm, SABRE, which can be applied to any MDP setting given access to a blackbox PAC RL algorithm for that setting. SABRE applies concepts from active learning to reinforcement learning to provably control the number of queries to the safety oracle. SABRE works by iteratively exploring the state space to find regions where the agent is currently uncertain about safety. Our main theoretical results shows that, under appropriate technical assumptions, SABRE never takes unsafe actions during training, and is guaranteed to return a near-optimal safe policy with high probability. We provide a discussion of how our meta-algorithm may be applied to various settings studied in both theoretical and empirical frameworks.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 28 of its 50 references that are in the map. First posted on arXiv 26 Oct 2022.

Builds on

28 of 50 references.

References · 28

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2017Minimax Regret Bounds for Reinforcement Learningcited by 500
  3. 2017Safe Reinforcement Learning via Shieldingcited by 427
  4. 2017Safe Model-based Reinforcement Learning with Stability Guaranteescited by 347
  5. 2021Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learningcited by 337
  6. 2020Responsive Safety in Reinforcement Learning by PID Lagrangian Methodscited by 302
  7. 2018A Lyapunov-based Approach to Safe Reinforcement Learningcited by 278
  8. 2019End-to-End Safe Reinforcement Learning through Barrier Functions for Safety-Critical Continuous Control Taskscited by 256
  9. 2019Provably efficient RL with Rich Observations via Latent State Decodingcited by 169
  10. 2022A Review of Safe Reinforcement Learning: Methods, Theory and Applicationscited by 169
  11. 2019Lyapunov-based Safe Policy Optimization for Continuous Controlcited by 140
  12. 2020Provably Efficient Safe Exploration via Primal-Dual Policy Optimizationcited by 131
16 more
  1. 2021Representation Learning for Online and Offline RL in Low-rank MDPscited by 128
  2. 2019Kinematic State Abstraction and Provably Efficient Rich-Observation Reinforcement Learningcited by 114
  3. 2020Safe Reinforcement Learning in Constrained Markov Decision Processescited by 109
  4. 2020Learning to be Safe: Deep RL with a Safety Criticcited by 109
  5. 2017Trial without Error: Towards Safe Reinforcement Learning via Human Interventioncited by 102
  6. 2016Safe Exploration in Finite Markov Decision Processes with Gaussian Processescited by 72
  7. 2022Safe Reinforcement Learning by Imagining the Near Futurecited by 67
  8. 2021Safe Reinforcement Learning Using Advantage-Based Interventioncited by 45
  9. 2021Learning Barrier Certificates: Towards Safe Reinforcement Learning with Zero Training-time Violationscited by 36
  10. 2021Safe Reinforcement Learning with Linear Function Approximationcited by 33
  11. 2020Learning with Safety Constraints: Sample Complexity of Reinforcement Learning for Constrained MDPscited by 32
  12. 2021Safe Driving via Expert Guided Policy Optimizationcited by 29
  13. 2020SAMBA: Safe Model-Based & Active Reinforcement Learningcited by 28
  14. 2021Safe Pontryagin Differentiable Programmingcited by 25
  15. 2020Provably Safe PAC-MDP Exploration Using Analogiescited by 9
  16. 2022Safe Exploration Incurs Nearly No Additional Sample Complexity for Reward-free RLcited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 4; 2025: 1; 2026: 1.

7 papers.

  1. 2024A Policy Gradient Primal-Dual Algorithm for Constrained MDPs with Uniform PAC Guaranteescited by 5
  2. 2024Safety through feedback in Constrained RLcited by 5
  3. 2024Flipping-based Policy for Chance-Constrained Markov Decision Processescited by 3
  4. 2023Model-Assisted Probabilistic Safe Adaptive Control With Meta-Bayesian Learningcited by 1
  5. 2024A Hormetic Approach to the Value-Loading Problem: Preventing the Paperclip Apocalypse?cited by 0
  6. 2025Approximating Safety Feedback Without a Safety Oracle via Model Predictive Controlcited by 0
  7. 2026Safe Reinforcement Learning with Preference-based Constraint Inferencecited by 0