mlsanityData to 14 Oct 2026

Verifiable Reinforcement Learning via Policy Extraction

arXiv 1805.08328 ↗First posted 22 May 2018cs.LG stat.MLcited by 177

Not on the map. Why?

Cited by 54 in Deep reinforcement learning methods, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

While deep reinforcement learning has successfully solved many challenging control tasks, its real-world applicability has been limited by the inability to ensure the safety of learned policies. We propose an approach to verifiable reinforcement learning by training decision tree policies, which can represent complex policies (since they are nonparametric), yet can be efficiently verified using existing techniques (since they are highly structured). The challenge is that decision tree policies are difficult to train. We propose VIPER, an algorithm that combines ideas from model compression and imitation learning to learn decision tree policies guided by a DNN policy (called the oracle) and its Q-function, and show that it substantially outperforms two baselines. We use VIPER to (i) learn a provably robust decision tree policy for a variant of Atari Pong with a symbolic state space, (ii) learn a decision tree policy for a toy game based on Pong that provably never loses, and (iii) learn a provably stable decision tree policy for cart-pole. In each case, the decision tree policy achieves performance equal to that of the original DNN policy.

Key facts. Cited by 177 papers in the map (data through 14 Oct 2026). First posted on arXiv 22 May 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2018: 1; 2019: 2; 2020: 0; 2021: 26; 2022: 26; 2023: 33; 2024: 34; 2025: 23; 2026: 32.

177 papers.

Sort:

By area:

  1. 2018Scalable agent alignment via reward modeling: a research directioncited by 447
  2. 2021Interpretable Machine Learning: Fundamental Principles and 10 Grand Challengescited by 367
  3. 2022Multi-Game Decision Transformerscited by 214
  4. 2022A Review of Safe Reinforcement Learning: Methods, Theory and Applicationscited by 181
  5. 2021A Survey on Interpretable Reinforcement Learningcited by 73
  6. 2021Learning to Synthesize Programs as Interpretable and Generalizable Policiescited by 59
  7. 2021Provable Repair of Deep Neural Networkscited by 42
  8. 2023Safe Reinforcement Learning via Probabilistic Logic Shieldscited by 42
  9. 2023Bridging the Human-AI Knowledge Gap: Concept Discovery and Transfer in AlphaZerocited by 42
  10. 2024Generating Code World Models with Large Language Models Guided by Monte Carlo Tree Searchcited by 40
  11. 2022On Transforming Reinforcement Learning by Transformer: The Development Trajectorycited by 36
  12. 2022Provably Safe Reinforcement Learning: Conceptual Analysis, Survey, and Benchmarkingcited by 34
38 more
  1. 2021A Review of Formal Methods applied to Machine Learningcited by 33
  2. 2023Explainable Deep Reinforcement Learning: State of the Art and Challengescited by 33
  3. 2024RL-GPT: Integrating Reinforcement Learning and Code-as-policycited by 32
  4. 2021Iterative Bounding MDPs: Learning Interpretable Policies via Non-Interpretable Methodscited by 31
  5. 2022A Survey of Explainable Reinforcement Learningcited by 31
  6. 2024HardNet: Hard-Constrained Neural Networks with Universal Approximation Guaranteescited by 31
  7. 2023Neurosymbolic Reinforcement Learning and Planning: A Surveycited by 24
  8. 2024Interpretable Concept Bottlenecks to Align Reinforcement Learning Agentscited by 23
  9. 2021ProTo: Program-Guided Transformer for Program-Guided Taskscited by 21
  10. 2023Hierarchical Programmatic Reinforcement Learning via Learning to Compose Programscited by 20
  11. 2024Interpretable and Editable Programmatic Tree Policies for Reinforcement Learningcited by 18
  12. 2022GALOIS: Boosting Deep Reinforcement Learning via Generalizable Logic Synthesiscited by 16
  13. 2024Dynamic Model Predictive Shielding for Provably Safe Reinforcement Learningcited by 15
  14. 2021Neurosymbolic Transformers for Multi-Agent Communicationcited by 13
  15. 2024Synthesizing Programmatic Reinforcement Learning Policies with Large Language Model Guided Searchcited by 13
  16. 2019Mo"ET: Mixture of Expert Trees and its Application to Verifiable Reinforcement Learningcited by 12
  17. 2021Improving Human Sequential Decision-Making with Reinforcement Learningcited by 12
  18. 2022MAVIPER: Learning Decision Tree Policies for Interpretable Multi-Agent Reinforcement Learningcited by 12
  19. 2022Learning Interpretable, High-Performing Policies for Autonomous Drivingcited by 11
  20. 2022What can we Learn Even From the Weakest? Learning Sketches for Programmatic Strategiescited by 11
  21. 2021SyReNN: A Tool for Analyzing Deep Neural Networkscited by 10
  22. 2022Explainable Action Advising for Multi-Agent Reinforcement Learningcited by 10
  23. 2023Inverse Decision Modeling: Learning Interpretable Representations of Behaviorcited by 10
  24. 2023Efficient Symbolic Policy Learning with Differentiable Symbolic Expressioncited by 10
  25. 2024A Survey on Verification and Validation, Testing and Evaluations of Neurosymbolic Artificial Intelligencecited by 10
  26. 2019From explanation to synthesis: Compositional program induction for learning from demonstrationcited by 9
  27. 2021Learning Symbolic Rules for Interpretable Deep Reinforcement Learningcited by 9
  28. 2022Safe Neurosymbolic Learning with Differentiable Symbolic Executioncited by 9
  29. 2022Symbolic Visual Reinforcement Learning: A Scalable Framework with Object-Level Abstraction and Differentiable Expression Searchcited by 9
  30. 2023Learning to Identify Critical States for Reinforcement Learning from Videoscited by 9
  31. 2021Program Synthesis Guided Reinforcement Learning for Partially Observed Environmentscited by 8
  32. 2021Unsupervised Learning of Neurosymbolic Encoderscited by 8
  33. 2023Optimal Decision Tree Policies for Markov Decision Processescited by 8
  34. 2023Choosing Well Your Opponents: How to Guide the Synthesis of Programmatic Strategiescited by 8
  35. 2024Reinforcement Learning Gradients as Vitamin for Online Finetuning Decision Transformerscited by 8
  36. 2025Multimodal LLM-assisted Evolutionary Search for Programmatic Control Policiescited by 8
  37. 2025Learning Game-Playing Agents with Generative Code Optimizationcited by 8
  38. 2022On the Safety of Interpretable Machine Learning: A Maximum Deviation Approachcited by 7

50 most cited of 177 All 177 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2020CDT: Cascading Decision Trees for Explainable Reinforcement Learning
  2. 2026A Survey on the Verification of Reinforcement Learning Policies
  3. 2024Putting the Iterative Training of Decision Trees to the Test on a Real-World Robotic Task
  4. 2025TreeRL: LLM Reinforcement Learning with On-Policy Tree Search
  5. 2024Unveiling Options with Neural Decomposition
  6. 2019Hybrid Reinforcement Learning with Expert State Sequences
  7. 2026SPOTting the Future: Lookahead Explanations for Deep Reinforcement Learning
  8. 2017Thinking Fast and Slow with Deep Learning and Tree Search