frontier models
The Generated Video Chose Its Next Step From What It Had Actually Made
WorldGuide couples a planner and executor in visual world space, deciding actions from generated progress rather than a fixed script.
Summary
WorldGuide couples a planner and executor in visual world space, deciding actions from generated progress rather than a fixed script.
WorldGuide treats procedural video generation as closed-loop execution. From an initial image and a goal, its planner selects an atomic action, its executor generates the corresponding clip, and the next decision uses the generated result or stops when the task appears complete. A hierarchical memory keeps long tasks bounded, while a new dataset supplies about 59,000 step annotations across 245 tasks. The authors report 33.33% task success on WorldGuide-Bench versus 29.90% for a comparison model given reference plans, and 47.69% versus 32.73% on VideoCraft-Bench under goal-only conditioning. These remain benchmark videos, not verified physical task execution.
Why it matters
WorldGuide couples a planner and executor in visual world space, deciding actions from generated progress rather than a fixed script.
Limits and context
- The authors report 33.33% task success on WorldGuide-Bench versus 29.90% for a comparison model given reference plans, and 47.69% versus 32.73% on VideoCraft-Bench under goal-only conditioning.
- These remain benchmark videos, not verified physical task execution.
Key claims
WorldGuide couples a planner and executor in visual world space, deciding actions from generated progress rather than a fixed script.
Qualification: The authors report 33.33% task success on WorldGuide-Bench versus 29.90% for a comparison model given reference plans, and 47.69% versus 32.73% on VideoCraft-Bench under goal-only conditioning.
Evidence: source-2026-10-10-013
Sources
- arXiv preprint 2610.12459, original v1arXiv / submitting research authors · primary research
Corrections
No corrections have been recorded for this story.