G2A: From Generative Models to Robot Actions

Large-scale generative models pretrained on video, image, and 3D data learn rich representations of objects, spatial structure, physical interactions, and how the visual world changes over time. These models are a promising source of prior knowledge for robot learning, especially because large-scale robot data remains expensive to collect and limited in diversity.

However, most pretraining data contains no robot action labels, while a policy must ultimately produce executable actions that respect the embodiment, dynamics, and sensing constraints of a physical system. This mismatch creates an action gap between generative pretraining and robot policy learning.

A video model can predict how a scene should change, and a 3D generative model can capture the geometry of an environment — but neither directly specifies which robot actions will reliably produce the desired change. G2A places the interface between generative models and robot policies at the center, asking how models trained to represent or generate changes in the world can help robots learn the actions that cause those changes.

Leading researchers across generative world models, robot policy learning, embodiment grounding, human-video learning, and evaluation.

We invite submissions on any aspect of turning pretrained generative models into robot policies. We welcome new, preliminary, and in-progress work — including negative results, empirical comparisons, position papers, benchmarks, and evaluation methodologies.

Topics of interest

• Latent action learning. Inverse dynamics, action-conditioned generation, and latent action spaces learned from action-free video that link visual change to control.
• Object- and interaction-centric models. Representations centered on objects, contact, affordances, and interaction rather than raw pixels, and how they expose actionable structure.
• World and action models. Predictive models of dynamics turned into signals for planning, model-predictive control, and policy learning.
• Cross-embodiment transfer. Grounding disembodied knowledge in the kinematic, dynamic, and sensing constraints of robots so that what is learned stays physically realizable.
• Heterogeneous data sources. Policy learning that combines human video, simulation, and robot demonstrations, including alignment and curation across sources.
• From open-loop generation to closed-loop control. Turning generative rollouts into policies capable of execution, correction, and recovery when predictions diverge from real observations.
• Robot data efficiency. Action-free and action-light policy learning, and which forms of supervision give the greatest value per unit of collection cost.
• Evaluation. Benchmarks and protocols measuring action feasibility, controllability, robot-data efficiency, generalization, robustness to distribution shift, and inference cost.

Submission tracks

Submission guidelines

Submissions should follow the CoRL 2026 LaTeX template and be between 4 and 9 pages, excluding references and appendices. All papers must be submitted as anonymized PDFs for double-blind review via OpenReview (submission portal to be announced), with author names, affiliations, and acknowledgments removed and prior work cited in the third person.

Accepted papers will be featured in a poster session, with a small number selected for oral presentation. Proceedings are non-archival, allowing future conference or journal submissions; work under review at or accepted by other venues is welcome, though CoRL 2026 main-conference papers are not eligible.

Important dates

All deadlines are 11:59 PM Anywhere on Earth. Dates are tentative until the OpenReview portal opens.

Awards

Thanks to the generous sponsorship of Lambda, the workshop will present:

• Best Paper Award: $3,000 in compute credits, selected from all accepted papers across both tracks and announced during the closing synthesis.
• Oral Paper Awards: $1,500 in compute credits each, for a small number of accepted papers selected for oral presentation at the workshop, chosen on review scores and breadth of interest to the audience.
• All other accepted papers: $400 in compute credits.

An interactive half-day event rather than a sequence of talks. Invited talks act as focused inputs for the breakout problem-solving sessions that follow.

Tentative morning schedule. Times subject to change as the program is finalized.