benchmarks evals
Static Scenes Were Easier Than Moving Worlds
4DCodeBench asks agents to turn video into executable graphics programs for deformation, fluid flow and fracture.
Summary
4DCodeBench asks agents to turn video into executable graphics programs for deformation, fluid flow and fracture.
4DCodeBench evaluates whether agents can reconstruct dynamic scenes from video as programs that encode both structure and physical evolution. Its real and synthetic scenes include deformation, fluids and fracture. The authors report that frontier models' strong static reconstruction did not translate into reliable complex dynamics, making executable behavior—not visual resemblance alone—the benchmark target.
Why it matters
4DCodeBench asks agents to turn video into executable graphics programs for deformation, fluid flow and fracture.
Limits and context
- The authors report that frontier models' strong static reconstruction did not translate into reliable complex dynamics, making executable behavior—not visual resemblance alone—the benchmark target.
Key claims
4DCodeBench asks agents to turn video into executable graphics programs for deformation, fluid flow and fracture.
Qualification: The authors report that frontier models' strong static reconstruction did not translate into reliable complex dynamics, making executable behavior—not visual resemblance alone—the benchmark target.
Evidence: source-2026-10-05-003
Sources
- arXiv preprint 2610.03715arXiv · primary research
Corrections
No corrections have been recorded for this story.