9th Annual Conference on Robot Learning

The 9th Annual Conference on Robot Learning (CoRL 2025) is the leading venue for research on how robots can learn from data, experience, demonstrations, and interaction rather than relying solely on hand-engineered programs. Topics include imitation learning, reinforcement learning for robotic manipulation and locomotion, learning from human feedback, sim-to-real transfer, foundation models for robotics, multi-task and language-conditioned robot learning, and the evaluation of learned robot policies in realistic settings.

CoRL attracts roboticists, machine learning researchers, and cognitive scientists who share the conviction that learning is the path to generalizable robot capabilities that transfer across tasks and environments. The conference has grown rapidly in prominence as large-scale robot learning has become technically feasible, attracting both academic groups training robots in research labs and industry teams deploying learned manipulation and navigation policies in warehouses, homes, and healthcare settings.