World Modeling Workshop

Date: 4-6 February 2026
Location: Agora, Mila - Quebec AI Institute, 6666 Rue Saint-Urbain, Montréal, Canada
Online: Public webinar (freely available to all) (link TBA)

A fundamental desideratum of AI agents is the ability to model environment dynamics and transitions in response to both their own actions and external control signals. This capability, commonly referred to as world modeling (WM), is essential for prediction, planning, and generalization. Learning world models using deep learning has been an active area of research for nearly a decade. In recent years, the field has witnessed significant breakthroughs driven by advances in deep neural architectures and scalable learning paradigms. Multiple subfields, including self-supervised learning (SSL), generative modeling, reinforcement learning (RL), robotics, and large language models (LLMs), have tackled aspects of world modeling, often with different tools and methodologies. While these communities address overlapping challenges, they frequently operate in isolation. As a result, insights and progress in one area may go unnoticed in another, limiting opportunities for synthesis and collaboration. This workshop aims to bridge this gap between subfields of world modeling by fostering open dialogue, critical discussion, and cross-disciplinary exchange. By bringing together researchers from diverse backgrounds, from early-career researchers to established experts, we hope to establish a shared vocabulary, identify common challenges, and surface synergies that can move the field of world modeling forward.

Apply to Attend World Modeling Workshop

If you are interested to attend the workshop, you can apply here.

If your application is accepted, you will receive an email with the registration details. Priority for registration will be given to applicants with accepted papers.

Student/Post-doc registration fee: 180 CAD

Non-Student/Industry registration fee: 280 CAD

Call for Papers World Modeling Workshop

Participants interested in presenting a poster are invited to submit an extended abstract with maximum length of 4 pages, excluding references and appendix. Authors can use this opportunity to present early-stage results, insightful negative findings, opinion pieces, or novel datasets. All submissions will undergo a double-blinded peer-review process. The accepted papers are non-archival.

Please note that at least one of the authors of each submitting paper must register as a reviewer (with the paper submission form). If you are interested in joining the program committee as a reviewer, you can also sign up here!

Please make sure to use the ICLR LaTeX template. For camera ready submissions please refer to the section below (TBA).

Topics of Interest

A non-exhaustive list of relevant topics are:

Self-supervised representations for world modeling
Model-based RL and planning
Generative world modeling
Leveraging world knowledge in LLMs
Embodied AI and Robotics
Causality
World models for scientific discovery
Physical plausibility of world models
World models and AI safety
Datasets and benchmarks for evaluating world models

⚠️ Important Dates 📅

Abstract submission deadline: October 10, 2025 AoE
Paper submission deadline: October 17, 2025 AoE (Submit on OpenReview)
Final decisions to authors: November 17, 2025
Camera Ready deadline: TBA

Speakers World Modeling Workshop

(stay tuned for more confirmed speakers!)

Keynote

These talks offer broad perspectives on the field of world modeling. They reflect on long-term directions, emerging opportunities, and the potential of world models for advancing AI research over the coming decade.

Flatiron Institute, NYU

Methods Deep Dive

These talks are in-depth examinations of state-of-the-art methods for world modeling. They focus on the subtle algorithmic choices and practical implementation details that critically influence the quality of world models.

Tutorials

These tutorials are hands-on sessions that offer practical design choices, guide participants through implementation, and provide background on core topics and ingredients of world modeling.

Randall Balestriero
(Brown University, Meta)

Self-Supervised Learning

Ricky Chen (Meta)

Diffusion & Flow Matching

Alessandro Lazaric (Meta)

Reinforcement Learning

Schedule (times in EST) (Tentative) World Modeling Workshop

Time	Day 1 – Wednesday, 04-02-2026	Day 2 – Thursday, 05-02-2026	Day 3 – Friday, 06-02-2026
08:30	Registration and Breakfast	Breakfast	Breakfast
09:00	Opening Remarks	Keynote 6	Tutorial 2
09:15	Keynote 1	Keynote 6	Tutorial 2
10:00	Keynote 2	Tutorial 1	Tutorial 2
10:30	Keynote 2	Tutorial 1	Coffee Break
10:45	Coffee Break	Tutorial 1	Tutorial 3
11:00	Keynote 3	Coffee Break	Tutorial 3
11:15	Keynote 3	Tutorial 1	Tutorial 3
11:45	Keynote 4	Methods Deep Dive	Tutorial 3
12:30	Lunch	Lunch	Lunch
13:30	Keynote 5	Poster Session 1	Poster Session 2
14:30	Panel Discussion	Poster Session 1	Poster Session 2
15:30	Methods Deep Dive	Poster Session 1	Coffee Break
15:45	Methods Deep Dive	Methods Deep Dive	Methods Deep Dive
16:00	Coffee Break	Methods Deep Dive	Methods Deep Dive
16:15	Lightning Talks 1	Methods Deep Dive	Methods Deep Dive
17:30	Reception (tentative)	Informal Networking / Overflow	Closing Remarks