Interactive Imitation Learning

Follow this topic Ask about this topic

Interactive imitation learning studies how an agent can improve by asking an expert for demonstrations or corrections on states it encounters, rather than learning only from a fixed dataset. Methods such as DAgger collect labels on learner-visited states, while selective intervention, shared control, and retrieval-based approaches aim to make each correction more useful and reduce expert effort. The field is applied to settings including robotic manipulation, navigation, and vision-language-action policies, with evaluation often measuring task success, recovery from errors, and human time. A central tension is when online interaction justifies its cost: it can help when the learner cannot represent the expert or faces unfamiliar states, but corrections may be costly, delayed, or poorly matched to the learner.

Latest papers

50 papers11 papers in the last four weeks, up 57% on the four weeks before. 0.1% of all new papers.

SimPT: An Automated Simulation Framework for Intervention-Based VLA Post-Training
Oct 21, 2026Caicheng Wang, Zipeng Ye, Zhexuan Zhou +3VLA Model Post-Training Vision-Language-Action Models
Meituan
TILDE: Teacher-in-the-Loop Distillation from Embodied Foundation Models
Oct 18, 2026Yuxiang Lu, Yantong Liu, Jinghua Hou +4Knowledge Distillation Interactive Imitation Learning
The University of Hong Kong
EqC: Equilibrium Energy Control for Introspective Lifelong Robot Learning
Oct 13, 2026Haodong Yang, Runhan Huang, Bowen Zhang +7Robot Policy Adaptation Continual Robot Learning
Tsinghua University · Galaxea AI · Princeton University
I3L: Iterative Interactive Imitation Learning for Whole-Body Mobile Manipulation
Oct 13, 2026Wensi Ai, Christopher Luey, Karthik Dharmarajan +7Robot Imitation Learning Interactive Imitation Learning
Stanford University · Lambda Inc.
Improving Policy Learning from Delayed Interventions
Oct 12, 2026Xiru Zhu, David Meger, Gregory DudekImitation Learning Robot Imitation Learning
McGill University, Computer Science
RoboPrompt: Intuitive Robot Policy Steering with Sparse Human Input
Oct 7, 2026Yanwen Zou, Chenyang Shi, Guoxuan Xu +5Robot Policy Adaptation Interactive Imitation Learning
Shanghai Jiao Tong University · Shanghai Innovation Institute · Noematrix Ltd.
ReDex: Repairing Sim-to-Real Dexterous Policies by Finger-Level Compliant Interaction
Oct 5, 2026Jinzhou Li, Hadi Tabatabaee, Kelin Yu +6Sim-to-Real Transfer Robotic Manipulation
Amazon · Duke University · University of Maryland, College Park +2
Mulligan: Performance-Guided Data Collection for Efficient On-Robot Learning
Oct 5, 2026Lars Ankile, Perry Dong, Rohan Bhowmik +4Robotic Data Collection Robotic Manipulation
Stanford University
Reward-DAgger: Robot-Gated Interactive Imitation Learning with General-Purpose Progress-Based Reward Models
Oct 2, 2026Ryan Li, Yigit Korkmaz, Erdem BıyıkRobot Imitation Learning Interactive Imitation Learning
Department of Mathematics, University of Southern California · Thomas Lord Department of Computer Science, University of Southern California
Plan-Conditioned Imitation for Robust Object Retrieval under Self-Occlusion in Dense Clutter
Sep 30, 2026Kowndinya Boyalakuntla, Ajinkya Pawar, Abdeslam Boularias +1Imitation Learning Robotic Manipulation
Rutgers University · Indian Institute of Technology Bombay, India
BlenDAgger: Blended Shared Control for Interactive Imitation Learning
Sep 29, 2026Cailyn Smith, Geoffrey Sun, Henny Admoni +1Robotic Control Robotic Manipulation
The Robotics Institute, School of Computer Science Carnegie Mellon University Pittsburgh, USA · School of Computer Science Carnegie Mellon University Pittsburgh, USA
RE-0: Verified Recursive Improvement of Embodied Code-as-Policy Agents through Local On-Policy Distillation
Sep 26, 2026Jiawei Zhang, Xiangrong Zhang, Rui Song +3Robot Policy Learning On-Policy Distillation
Jilin University, Changchun, China · Dalian University of Technology, Dalian, China
ForceRFT: Refining VLA Actions through Force-Guided Residual Reinforcement Learning
Sep 19, 2026Yichen Wang, Chaoyang Zhang, Xuqi Su +3Robotic Manipulation Vision-Language-Action Models
College of Information Science and Technology, Eastern Institute of Technology, Ningbo, Ningbo 315200, China. · Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment and Technology, Jiangnan University, Wuxi, Jiangsu, China. · School of Automation and Intelligent Sensing, Shanghai Jiao Tong University, Shanghai, China. +3
HIL-UMI: Bringing Human-in-the-Loop Post-Training of Vision-Language-Action Models to Universal Manipulation Interface
Sep 17, 2026Zimu Han, Yiming Zeng, Jiyao Zhang +9Interactive Imitation Learning Vision-Language-Action Models
Center on Frontier Computing Studies, School of Computer Science, Peking University, China · Xi’an Jiaotong University, China · National Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University, China +2
Towards High-DoF Dexterous Manipulation through VLA Post-Training
Sep 17, 2026Junlei Zhu, Shenzhe Yao, Chaogui Huang +6Latent Action Learning VLA Model Post-Training
Wuji Technology · School of Information Science and Technology, ShanghaiTech University, China
Missing Bridges: Composition-Aware Active Imitation Learning
Sep 16, 2026Maxwell J. Jacobson, Ahmed H Qureshi, Yexiang XueImitation Learning Interactive Imitation Learning
Department of Computer Science, Purdue University West Lafayette, Indiana, USA
Co-Evolving Harnesses and Models: On-Policy Correction Helps Weaker Models Catch Up Where Imitation Fails
Sep 8, 2026Zhou Yu, Bin Bi, Shiva Kumar Pentyala +8Agent Harness Optimization Imitation Learning
Salesforce AI
DISEIL: Demonstration Distillation for Sample-Efficient Imitation Learning
Sep 8, 2026Suyog Khanal, Arun Kumar A, Santu RanaImitation Learning Interactive Imitation Learning
Deakin Applied Artificial Intelligence Initiative Geelong, Australia
Self-Aware Active Learning Enables Continual Improvement in Autonomous Driving
Aug 30, 2026Dong Hu, Chao Huang, Carman K. M. Lee +1Active Learning Autonomous Driving
Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong · School of Electrical and Mechanical Engineering, Adelaide University, Australia · Department of Computer Science, University College London, United Kingdom
Max-Q Selective Imitation for Human-in-the-Loop Online Robot Learning
Aug 15, 2026Zihang Wang, Yishan WangRL for Robotics Reinforcement Learning
AutoIntervene: Calibrated Intervention for Action-Chunking Imitation Learning Policies
Aug 7, 2026Jinhe Tang, Weiming ZhiRobot Policy Learning Interactive Imitation Learning
School of Computer Science · Australian Center For Robotics · College of Connected Computing
When Does On-Policy Interaction Help? Representational Tradeoffs in Value-Based Imitation Learning
Jul 31, 2026Luca Viano, Antoine Moulin, Audrey Huang +3Imitation Learning Interactive Imitation Learning
Microsoft
Never Too Late for Force: Accelerating VLA Post-Training with Reactive Force Injection
Jul 15, 2026Yi Wang, Wendi Chen, Zimo Wen +8Robotic Control Language-Conditioned Robot Manipulation
Shanghai Jiao Tong University · Shanghai Innovation Institute · Zhiyuan College, Shanghai Jiao Tong University +2
Joint On-and-Off Policy Learning for Vision-and-Language Navigation
Jul 15, 2026Qingrong He, Lin Zhao, Kevin Zheng +1Vision-Language Navigation Interactive Imitation Learning
All authors are with Joy Future Academy.
FlowDAgger: Human-in-the-Loop Adaptation of Generative Robot Policies in Latent Space
Jul 9, 2026Michael Murray, Daphne Chen, Simran Bagaria +7Latent Action Learning Robot Policy Learning
1Microsoft Research · University of Washington · 2Microsoft +1
When Does Online Imitation Learning Help in LLM Post-Training? The Role of (Non-)Realizability Beyond Horizon
Jun 29, 2026Huaqing Zhang, Jingchu Gai, Juno Kim +2Imitation Learning Interactive Imitation Learning
Tsinghua University · Carnegie Mellon University · UC Berkeley +1
World Action Models Enable Continual Imitation Learning with Recurrent Generative Replays
Jun 25, 2026Manish Kumar Govind, Dominick Reilly, Smit Patel +2Generative Replay Continual Robot Learning
Department of Computer Science University of North Carolina at Charlotte, United States
RE4: Transformation-aware Imitation of Object Interactions Using Manipulation Modes
Jun 23, 2026Arsh Chawla, Rahul ShomeImitation Learning Robotic Manipulation
Australian National University
ARP: Enhancing Quantized Skill Abstractions via Visual Alignment and Iterative Refinement for Robotic Manipulation
Jun 21, 2026Yuntian Wang, Zesheng Jia, Yuhui Duan +5Long-Horizon Robotic Manipulation Robot Skill Learning
School of Future Science and Engineering, Soochow University · Key Laboratory of General Artificial Intelligence and Large Models in Provincial Universities, Soochow University · College of Mechanical and Electronic Engineering, China Jiliang University +2
Bidirectional Tutoring for Developmental Motor Learning in Robots: Co-Developed Interaction Dynamics Support Stable Learning
Jun 18, 2026Rui Fukushima, Jun TaniGenerative Replay Robot Skill Learning
Cognitive Neurorobotics Research Unit, Okinawa Institute of Science and Technology Graduate University, Okinawa, Japan

Oct 21, 2026J/K move · Enter open · S save

SimPT: An Automated Simulation Framework for Intervention-Based VLA Post-Training

Caicheng Wang, Zipeng Ye, Zhexuan Zhou, Zilong Huang+2

Meituan

VLA Model Post-Training Vision-Language-Action Models Interactive Imitation Learning Robotic Manipulation Robot Imitation Learning

Open paper AskPDF arXiv

Human-in-the-loop (HIL) post-training improves vision-language-action (VLA) policies through expert corrections, but costly human supervision and repeated real-robot evaluation limit algorithm exploration and development. We introduce SimPT (Simulation-based Post-Training), a simulation framework that automates intervention-based VLA post-training by integrating policy rollouts, corrective data collection, policy training, and standalone evaluation. To automatically determine intervention timing and recovery strategies, we develop an Agent-Compiled Progress Graph that tracks task progress using privileged simulator state and retains stable snapshots during policy execution. After a rollout fails, a CuRobo-based expert attempts recovery from these snapshots, starting at more advanced task stages and falling back to earlier ones as needed to generate corrective trajectories. We implement and compare four post-training methods, RECAP, HG-DAgger, DPO+SFT, and FlowPRO, across seven RoboTwin manipulation tasks. Relative to a fixed-interval snapshot baseline, the Progress Graph reduces recovery attempts by 75.0% and increases recovery success from 93.4% to 98.9%. Using SimPT, we develop RECAP-CF-LPF, a refinement of RECAP that combines cross-fitting and low-pass filtering of value predictions. This refinement increases the overall success rate from 76.50% to 79.36% in held-out simulation trials. In real-robot HIL post-training, it increases the success rate from 72% to 96% across 25 scenarios of an Ethernet cable-insertion task. These results demonstrate the utility of SimPT for exploring and developing VLA post-training algorithms.