From passive generation to interactive agents that strategically decide what to sense, when to sense, and how to act.
Generative models have mastered passive generation. But real intelligence is active. It observes, plans, acts, and learns from feedback.
Models that strategically choose what to observe—optimal viewpoints, sensor placement, information seeking.
Continuous replanning based on new observations. Perception and action form a tight feedback loop.
Agents that learn through interaction. Active decision-making transforms passive models into interactive systems.
Leading voices in vision, robotics, and embodied AI
University of Maryland
Stanford University
Stanford University
UC Berkeley
MIT CSAIL
* listed alphabetically by last name
Opening Welcome & Introductions
10 min
Invited Talk: Speaker 1
30 min
Invited Talk: Speaker 2
30 min
☕ Coffee Break
15 min
Invited Talk: Speaker 3
30 min
Invited Talk: Speaker 4
30 min
Invited Talk: Speaker 5
30 min
Panel Discussion
Ranjay Krishna (UW & AI2), Amir Bar (Meta), Yilun Du (Harvard)
Closing Remarks
Submit 4-page short or 8-page long papers on world models, active sensing, embodied planning, and related topics. Non-archival, so work in progress is welcome.
Deadline: March 27, 2026
Notifications: April 15, 2026
Benchmark your world models on active embodied tasks. Four tracks: Active Recognition, Embodied QA, Image-Goal Navigation, Robotic Manipulation.
Deadline: May 28, 2026
Awards: Compute credits, best paper prizes
listed alphabetically by last name
Rama Chellappa
JHU
Jieneng Chen
JHU
Contact Person
Yilun Du
Harvard
Sanjeev Khudanpur
JHU
Cheng Peng
University of Virginia
Tianmin Shu
JHU
Chen Wei
Rice University
Jianwen Xie
Lambda
Alan Yuille
JHU
Questions about the workshop, submissions, or anything else? Reach out to our contact person.
Jieneng Chen • jchen293@jh.edu