Machine Learning/CORE Unranked

R2RL 2026

CoRL 2026 Workshop on Roadmap to Sample-Efficient Real-World Reinforcement Learning

1794459600000Austin, TXOfficial workshop site Site reachable

The R2RL 2026 workshop focuses on overcoming the sample efficiency bottleneck in real-world reinforcement learning for robotics, bringing together early-career researchers to identify practical challenges and co-create a roadmap through interactive sessions. The workshop features crowdsourced problem boards, breakout discussions, and a post-workshop whitepaper to document findings and guide future research.

Official CFP Back to deadlines Verified September 9, 2026

Paper fit

Contribution paths

A strong submission should clearly identify its contribution and evaluate it appropriately.

Full papers

Submissions of 4–8 pages (excluding references and appendix) following the official CoRL paper template and style, anonymized for double-blind review. Accepted papers will be presented in a poster session, with a subset selected for 5-minute spotlight talks.

Research areas in scope

Topics of interest

Supervision and Reward Cost: Methods for obtaining useful learning signals cheaply, including learned reward models, human feedback, human-in-the-loop interventions, and strategies for avoiding supervision as the bottleneck.Exploration and Safety on Hardware: Algorithms and systems for efficient exploration under safety constraints, reset-free or autonomous RL, safe online adaptation, and approaches that improve wall-clock efficiency during real-world training.Leveraging Priors for Real-World RL: Approaches that use pretrained policies, VLAs, offline datasets, world models, or sim-to-real transfer to reduce the number of required on-robot interactions.Post-Deployment Adaptation: Methods for finetuning deployed policies, improving robustness to distribution shift, learning new behaviors from experience, and adapting generalist policies without catastrophic forgetting.Long-Horizon and Contact-Rich RL: Techniques for improving sample efficiency in long-horizon manipulation, dexterous control, contact-rich tasks, and settings with delayed rewards, compounding errors, or difficult exploration.Benchmarks, Metrics, and Lessons Learned: Benchmarks, evaluation protocols, shared platforms, negative results, system-level insights, and analysis of what worked, what failed, and why in real-world RL experiments.

Policies worth checking twice

Submissions must be anonymized for double-blind review.
Submissions must follow the official CoRL paper template and style.
Submissions must be 4–8 pages, excluding references and appendix.
Work already accepted to the main CoRL 2026 conference is not allowed.
Accepted workshop papers are non-archival and will not appear in formal proceedings.

Official sources

Compiled from the official call for papers. The organizers’ pages remain authoritative.

Official workshop site

Last verified September 9, 2026