mlsanityData to 7 Oct 2026

EnsembleDAgger: A Bayesian Approach to Safe Imitation Learning

arXiv 1807.08364 ↗First posted 22 Jul 2018cs.LG cs.AIcited by 96

On the map: Offline RL and imitation learning from data → Imitation learning methods

Abstract (arXiv)

While imitation learning is often used in robotics, the approach frequently suffers from data mismatch and compounding errors. DAgger is an iterative algorithm that addresses these issues by aggregating training data from both the expert and novice policies, but does not consider the impact of safety. We present a probabilistic extension to DAgger, which attempts to quantify the confidence of the novice policy as a proxy for safety. Our method, EnsembleDAgger, approximates a Gaussian Process using an ensemble of neural networks. Using the variance as a measure of confidence, we compute a decision rule that captures how much we doubt the novice, thus determining when it is safe to allow the novice to act. With this approach, we aim to maximize the novice's share of actions, while constraining the probability of failure. We demonstrate improved safety and learning performance compared to other DAgger variants and classic imitation learning on an inverted pendulum and in the MuJoCo HalfCheetah environment.

Key facts. Cited by 96 papers in the map (data through 7 Oct 2026). Placed in Imitation learning methods, one of 13 directions in Offline RL and imitation learning from data. Builds on 9 of its 29 references that are in the map. First posted on arXiv 22 Jul 2018.

Builds on

9 of 29 references.

References · 9

  1. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,416
  2. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,888
  3. 2016OpenAI Gymcited by 1,684
  4. 2016Concrete Problems in AI Safetycited by 1,582
  5. 2017Deep Neural Networks as Gaussian Processescited by 617
  6. 2016Benchmarking Deep Reinforcement Learning for Continuous Controlcited by 361
  7. 2017Safe Model-based Reinforcement Learning with Stability Guaranteescited by 347
  8. 2016Query-Efficient Imitation Learning for End-to-End Autonomous Drivingcited by 83
  9. 2016Safe Learning of Regions of Attraction for Uncertain, Nonlinear Systems with Gaussian Processescited by 38

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2018: 1; 2019: 1; 2020: 1; 2021: 8; 2022: 10; 2023: 17; 2024: 19; 2025: 19; 2026: 20.

96 papers.

Sort:

By area:

  1. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  2. 2018HG-DAgger: Interactive Imitation Learning with Human Expertscited by 245
  3. 2024RT-H: Action Hierarchies Using Languagecited by 204
  4. 2023A Survey of Imitation Learning: Algorithms, Recent Developments, and Challengescited by 145
  5. 2022Robot Learning on the Job: Human-in-the-Loop Autonomy and Learning During Deploymentcited by 101
  6. 2021ThriftyDAgger: Budget-Aware Novelty and Risk Gating for Interactive Imitation Learningcited by 98
  7. 2020Safe Imitation Learning via Fast Bayesian Reward Inference from Preferencescited by 73
  8. 2021Error-Aware Imitation Learning from Teleoperation Data for Mobile Manipulationcited by 59
  9. 2023NeRF in the Palm of Your Hand: Corrective Augmentation for Robotics via Novel-View Synthesiscited by 56
  10. 2021LazyDAgger: Reducing Context Switching in Interactive Imitation Learningcited by 50
  11. 2024Bidirectional Decoding: Improving Action Chunking via Guided Test-Time Samplingcited by 46
  12. 2021On the Sample Complexity of Stability Constrained Imitation Learningcited by 38
38 more
  1. 2023Model-Based Runtime Monitoring with Interactive Imitation Learningcited by 37
  2. 2025Failure Prediction at Runtime for Generative Robot Policiescited by 36
  3. 2024Multi-Task Interactive Robot Fleet Learning with Visual World Modelscited by 33
  4. 2022Fleet-DAgger: Interactive Robot Fleet Learning with Scalable Human Supervisioncited by 29
  5. 2024Diff-DAgger: Uncertainty Estimation with Diffusion Policy for Robotic Manipulationcited by 29
  6. 2023Human-in-the-Loop Task and Motion Planning for Imitation Learningcited by 28
  7. 2023RLIF: Interactive Imitation Learning as Reinforcement Learningcited by 24
  8. 2022Eliciting Compatible Demonstrations for Multi-Human Imitation Learningcited by 22
  9. 2022When to Ask for Help: Proactive Interventions in Autonomous Reinforcement Learningcited by 20
  10. 2022PATO: Policy Assisted TeleOperation for Scalable Robot Data Collectioncited by 19
  11. 2021Safe Reinforcement Learning with Chance-constrained Model Predictive Controlcited by 18
  12. 2026When to Trust Imagination: Adaptive Action Execution for World Action Modelscited by 18
  13. 2023MEGA-DAgger: Imitation Learning with Multiple Imperfect Expertscited by 17
  14. 2024So You Think You Can Scale Up Autonomous Robot Data Collection?cited by 17
  15. 2023Safety-aware Causal Representation for Trustworthy Offline Reinforcement Learning in Autonomous Drivingcited by 16
  16. 2025Learning from Active Human Involvement through Proxy Value Propagationcited by 16
  17. 2022State Dropout-Based Curriculum Reinforcement Learning for Self-Driving at Unsignalized Intersectionscited by 15
  18. 2023Asking for Help: Failure Prediction in Behavioral Cloning through Value Approximationcited by 15
  19. 2024SAFE-GIL: SAFEty Guided Imitation Learning for Robotic Systemscited by 13
  20. 2023Active Reward Learning from Online Preferencescited by 12
  21. 2023A Survey of Demonstration Learningcited by 12
  22. 2024Conformalized Interactive Imitation Learning: Handling Expert Shift and Intermittent Feedbackcited by 11
  23. 2024VIEW: Visual Imitation Learning with Waypointscited by 10
  24. 2023How To Guide Your Learner: Imitation Learning with Active Adaptive Expert Involvementcited by 8
  25. 2024Efficient Active Imitation Learning with Random Network Distillationcited by 7
  26. 2025MILE: Model-based Intervention Learningcited by 7
  27. 2025Genie Centurion: Accelerating Scalable Real-World Robot Training with Human Rewind-and-Refine Guidancecited by 7
  28. 2026ALOE: Action-Level Off-Policy Evaluation for Vision-Language-Action Model Post-Trainingcited by 7
  29. 2025Robot-Gated Interactive Imitation Learning with Adaptive Intervention Mechanismcited by 6
  30. 2026Hide-and-Seek in Trajectories: Discovering Failure Signals for VLA Runtime Monitoringcited by 6
  31. 2019Better-than-Demonstrator Imitation Learning via Automatically-Ranked Demonstrationscited by 5
  32. 2024Leveraging Demonstrator-perceived Precision for Safe Interactive Imitation Learning of Clearance-limited Taskscited by 5
  33. 2024RAIL: Reachability-Aided Imitation Learning for Safe Policy Executioncited by 5
  34. 2024Contractive Dynamical Imitation Policies for Efficient Out-of-Sample Recoverycited by 5
  35. 2025Real-Time Operator Takeover for Visuomotor Diffusion Policy Trainingcited by 5
  36. 2026DexHiL: A Human-in-the-Loop Framework for Vision-Language-Action Model Post-Training in Dexterous Manipulationcited by 5
  37. 2021Fully General Online Imitation Learningcited by 4
  38. 2023IIFL: Implicit Interactive Fleet Learning from Heterogeneous Human Supervisorscited by 4

50 most cited of 96 All 96 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2022Imitation Learning by Estimating Expertise of Demonstrators
  2. 2026TAIL-Safe: Task-Agnostic Safety Monitoring for Imitation Learning Policies
  3. 2025RISE: Robust Imitation through Stochastic Encoding
  4. 2021Bayesian Disturbance Injection: Robust Imitation Learning of Flexible Policies
  5. 2021Confidence-Aware Imitation Learning from Demonstrations with Varying Optimality
  6. 2019Uncertainty-Aware Imitation Learning using Kernelized Movement Primitives
  7. 2026To Do or Not to Do: Ensuring the Safety of Visuomotor Policies Learned from Demonstrations
  8. 2022Bayesian Disturbance Injection: Robust Imitation Learning of Flexible Policies for Robot Manipulation