mlsanityData to 7 Oct 2026

Your Learned Constraint is Secretly a Backward Reachable Tube

arXiv 2501.15618 ↗First posted 26 Jan 2025cs.RO cs.AI cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

Inverse Constraint Learning (ICL) is the problem of inferring constraints from safe (i.e., constraint-satisfying) demonstrations. The hope is that these inferred constraints can then be used downstream to search for safe policies for new tasks and, potentially, under different dynamics. Our paper explores the question of what mathematical entity ICL recovers. Somewhat surprisingly, we show that both in theory and in practice, ICL recovers the set of states where failure is inevitable, rather than the set of states where failure has already happened. In the language of safe control, this means we recover a backwards reachable tube (BRT) rather than a failure set. In contrast to the failure set, the BRT depends on the dynamics of the data collection system. We discuss the implications of the dynamics-conditionedness of the recovered constraint on both the sample-efficiency of policy search and the transferability of learned constraints.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 25 of its 41 references that are in the map. First posted on arXiv 26 Jan 2025.

Builds on

25 of 41 references.

References · 25

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2016Generative Adversarial Imitation Learningcited by 1,576
  3. 2019Control Barrier Functions: Theory and Applicationscited by 332
  4. 2020Responsive Safety in Reinforcement Learning by PID Lagrangian Methodscited by 302
  5. 2017Inverse Reward Designcited by 246
  6. 2017Hamilton-Jacobi Reachability: A Brief Overview and Recent Advancescited by 186
  7. 2021Tianshou: a Highly Modularized Deep Reinforcement Learning Librarycited by 139
  8. 2020DeepReach: A Deep Learning Approach to High-Dimensional Reachabilitycited by 81
  9. 2021Of Moments and Matching: A Game-Theoretic Framework for Closing the Imitation Gapcited by 80
  10. 2014Reach-Avoid Problems with Time-Varying Dynamics, Targets and Constraintscited by 41
  11. 2023Inverse Reinforcement Learning without Reinforcement Learningcited by 35
  12. 2022Sequence Model Imitation Learning with Unobserved Contextscited by 34
13 more
  1. 2019Maximum Likelihood Constraint Inference for Inverse Reinforcement Learningcited by 33
  2. 2024Hybrid Inverse Reinforcement Learningcited by 31
  3. 2022ISAACS: Iterative Soft Adversarial Actor-Critic for Safetycited by 24
  4. 2023What Went Wrong? Closing the Sim-to-Real Gap via Differentiable Causal Discoverycited by 22
  5. 2023Learning Safety Constraints from Demonstrations with Unknown Rewardscited by 19
  6. 2024Diffusing States and Matching Scores: A New Framework for Imitation Learningcited by 16
  7. 2021Maximum Likelihood Constraint Inference from Stochastic Demonstrationscited by 15
  8. 2023Learning Shared Safety Constraints from Multi-task Demonstrationscited by 15
  9. 2024A Comprehensive Survey on Inverse Constrained Reinforcement Learning: Definitions, Progress and Challengescited by 10
  10. 2024EvIL: Evolution Strategies for Generalisable Imitation Learningcited by 7
  11. 2022CALIPSO: A Differentiable Solver for Trajectory Optimization with Conic and Complementarity Constraintscited by 6
  12. 2023Learning Covariances for Estimation with Constrained Bilevel Optimizationcited by 4
  13. 2023Learning Observation Models with Incremental Non-Differentiable Graph Optimizers in the Loop for Robotics State Estimationcited by 2

Built on by

Cited by year (2026 to 7 Oct)

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

2 papers.

  1. 2025Learning Neural Control Barrier Functions from Expert Demonstrations using Inverse Constraint Learningcited by 0
  2. 2026Multi-Objective Constraint Inference using Inverse reinforcement learningcited by 0