Overview

Imitation learning (IL) has rapidly become the dominant paradigm for training robot policies, powering recent advances in dexterous manipulation, household tasks, and large-scale foundation models. Despite this success, IL methods face persistent challenges in data efficiency, robustness, and generalization to unseen conditions.

Reinforcement learning (RL), with its emphasis on exploration and reward optimization, offers complementary strengths that could help address these limitations—but its integration with IL remains an open problem.

How can reinforcement learning improve the real-world performance of robot policies, especially when complementing imitation learning?

This workshop aims to convene leading researchers from academia and industry to examine challenges and opportunities at this intersection, including efficient fine-tuning of pretrained policies, uncertainty-aware and human-in-the-loop adaptation, the use of real-to-sim digital twins for continuous policy improvement, and benchmarks for real-world evaluation.

Important Dates

Submission Opens: Feb 10, 2026
Paper Submission Deadline: Mar 20, 2026
Notification of Acceptance: TBD
Camera Ready: TBD
Workshop Date: June 1, 2026

Schedule

Time	Session
08:45–09:00	Opening remarks (organizers)
09:00–09:30	Invited Talk: Sergey Levine
09:30–10:00	Invited Talk: Chelsea Finn
10:00–10:30	Lightning Talks I for Accepted Papers (3–5 minutes each)
10:30–11:00	Coffee break + Posters I
11:00–11:30	Invited Talk: Robert Platt
11:30–12:00	Invited Talk: Pulkit Agrawal
12:00–12:30	Panel Discussion: "Reinforce or Imitate: Practical Challenges in RL with Real World Data"
12:30–14:00	Lunch
14:00–14:30	Invited Talk: Jason Ma
14:30–15:00	Invited Talk: Georgia Chalvatzaki
15:00–15:30	Invited Talk: David Held
15:30–16:00	Lightning Talks II for Accepted Papers (3–5 minutes each)
16:00–16:30	Coffee break + Posters II
16:30–17:00	Award Announcement and Closing Remarks

Invited Speakers

Sergey Levine

UC Berkeley, Physical Intelligence

Chelsea Finn

Stanford University, Physical Intelligence

Pulkit Agrawal

MIT

Robert Platt

Northeastern University

Jason Ma

Dyna Robotics

Georgia Chalvatzaki

TU Darmstadt

David Held

CMU

Call for Papers

We invite submissions of research works in the following domains, including but not limited to:

Reinforcement Learning Fine-tuning
Reward learning / Preference Learning
Imitation Learning Based Policy Fine-tuning
Real2Sim for Continuous Policy Improvement
Benchmark Design for Real-World Evaluation

Submission Requirements

Format: Papers must follow the IEEE conference format. Use the official ICRA template available at https://www.ieee.org/conferences/publishing/templates.html
Page Limit: Submissions should not exceed 6 pages, excluding references
Review Process: All submissions will undergo a double-blind peer review process
Submission Portal: Papers should be submitted via OpenReview (link below)

Submit Paper (OpenReview)

Accepted papers will be presented through lightning talks and poster sessions, with opportunities for in-depth technical discussions.