mlsanityData to 7 Oct 2026

Probabilistic Reach-Avoid for Bayesian Neural Networks

arXiv 2310.01951 ↗First posted 3 Oct 2023cs.LG cs.AIcited by 4

Not on the map. Why?

Abstract (arXiv)

Model-based reinforcement learning seeks to simultaneously learn the dynamics of an unknown stochastic environment and synthesise an optimal policy for acting in it. Ensuring the safety and robustness of sequential decisions made through a policy in such an environment is a key challenge for policies intended for safety-critical scenarios. In this work, we investigate two complementary problems: first, computing reach-avoid probabilities for iterative predictions made with dynamical models, with dynamics described by Bayesian neural network (BNN); second, synthesising control policies that are optimal with respect to a given reach-avoid specification (reaching a "target" state, while avoiding a set of "unsafe" states) and a learned BNN model. Our solution leverages interval propagation and backward recursion techniques to compute lower bounds for the probability that a policy's sequence of actions leads to satisfying the reach-avoid specification. Such computed lower bounds provide safety certification for the given policy and BNN model. We then introduce control synthesis algorithms to derive policies maximizing said lower bounds on the safety probability. We demonstrate the effectiveness of our method on a series of control benchmarks characterized by learned BNN dynamics models. On our most challenging benchmark, compared to purely data-driven policies the optimal synthesis algorithm is able to provide more than a four-fold increase in the number of certifiable states and more than a three-fold increase in the average guaranteed reach-avoid probability.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 35 of its 66 references that are in the map. First posted on arXiv 3 Oct 2023.

Builds on

35 of 66 references.

References · 35

  1. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  2. 2019Mastering Atari, Go, Chess and Shogi by Planning with a Learned Modelcited by 1,327
  3. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,259
  4. 2017Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networkscited by 739
  5. 2017Safe Reinforcement Learning via Shieldingcited by 427
  6. 2017Safe Model-based Reinforcement Learning with Stability Guaranteescited by 347
  7. 2018On the Effectiveness of Interval Bound Propagation for Training Verifiably Robust Modelscited by 322
  8. 2021What Are Bayesian Neural Network Posteriors Really Like?cited by 307
  9. 2021The Bayesian Learning Rulecited by 95
  10. 2015Safe Controller Optimization for Quadrotors with Gaussian Processescited by 77
  11. 2018Adv-BNN: Improved Adversarial Defense through Robust Bayesian Neural Networkcited by 68
  12. 2019Uncertainty Quantification with Statistical Guarantees in End-to-End Autonomous Driving Controlcited by 60
23 more
  1. 2016Learning and Policy Search in Stochastic Dynamical Systems with Bayesian Neural Networkscited by 49
  2. 2020Robustness of Bayesian Neural Networks to Gradient-Based Attackscited by 48
  3. 2018Formal Verification of Neural Network Controlled Autonomous Systemscited by 41
  4. 2020Probabilistic Safety for Bayesian Neural Networkscited by 41
  5. 2019Certified Reinforcement Learning with Logic Guidancecited by 35
  6. 2019Statistical Guarantees for the Robustness of Bayesian Neural Networkscited by 35
  7. 2019Reinforcement Learning with Probabilistic Guarantees for Autonomous Drivingcited by 27
  8. 2019Neural Simplex Architecturecited by 27
  9. 2021Bayesian Inference with Certifiable Adversarial Robustnesscited by 26
  10. 2018Robustness Guarantees for Bayesian Inference with Gaussian Processescited by 23
  11. 2022Robust Explanation Constraints for Neural Networkscited by 23
  12. 2021Safe Control with Neural Network Dynamic Modelscited by 21
  13. 2021Make Sure You're Unsure: A Framework for Verifying Probabilistic Specificationscited by 18
  14. 2019Verifiably Safe Off-Model Reinforcement Learningcited by 16
  15. 2020Probabilistic Guarantees for Safe Deep Reinforcement Learningcited by 15
  16. 2021Infinite Time Horizon Safety of Bayesian Neural Networkscited by 12
  17. 2019Efficiency through Uncertainty: Scalable Formal Synthesis for Stochastic Hybrid Systemscited by 11
  18. 2019Adversarial Robustness Guarantees for Classification with Gaussian Processescited by 11
  19. 2017Safe Policy Search with Gaussian Process Modelscited by 9
  20. 2020Safety Verification of Unknown Dynamical Systems via Gaussian Process Regressioncited by 9
  21. 2021Certification of Iterative Predictions in Bayesian Neural Networkscited by 8
  22. 2022Formal Control Synthesis for Stochastic Neural Network Dynamic Modelscited by 7
  23. 2017How Wrong Am I? - Studying Adversarial Examples and their Impact on Uncertainty in Gaussian Process Machine Learning Modelscited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 1; 2026: 2.

4 papers.

  1. 2025Stochastic Weight Sharing for Bayesian Neural Networkscited by 2
  2. 2024Risk-Averse Certification of Bayesian Neural Networkscited by 0
  3. 2026Probabilistic Verification of Neural Networks via Efficient Probabilistic Hull Generationcited by 0
  4. 2026Hybrid Robustness Verification for Spatio-Temporal Neural Networkscited by 0