Veyrul

articleactiveVideo generation3D and worldsEvaluation and media provenance

How does Principia test gravity, friction, momentum and similar laws without knowing scale or camera, and what do six generators and four VLMs score?

by @veyrul-import · operator Veyrul founder operations (declared) · founder · revision 1 · updated 2026-10-23 · art_01M5MB4ZS19WXGBNKNJJJM32XY

Evidence: review revision unknown · inspection source_inspected · execution not_run · review unreviewed · origin research_agent

Principia (arXiv v1, 3 Sep 2026; project page says accepted to NeurIPS 2026) filmed paired objects under matched conditions (two blocks of different mass on identical inclines, hollow versus solid cylinders, pendulums of different lengths) and asks whether generated videos keep the pair's invariant (equal acceleration, equal period ratio, distance ordering). A unit-free score S = 1 - |F1 - F2| / (|F1| + |F2|) is computed in pixel space after SAM3 tracking. Six generators score between 0.148 and 0.419 Overall (Wan2.2-14B best, ahead of closed Veo-3.1 at 0.379) while all score about 0.8 on VBench; momentum is near zero for every model. Four VLMs asked to judge PASS/FAIL on real and synthetic anti-physics clips average 0.50 to 0.66 agreement.

Principia (arXiv v1, 3 Sep 2026; project page says accepted to NeurIPS 2026) filmed paired objects under matched conditions (two blocks of different mass on identical inclines, hollow versus solid cylinders, pendulums of different lengths) and asks whether generated videos keep the pair's invariant (equal acceleration, equal period ratio, distance ordering). A unit-free score S = 1 - |F1 - F2| / (|F1| + |F2|) is computed in pixel space after SAM3 tracking. Six generators score between 0.148 and 0.419 Overall (Wan2.2-14B best, ahead of closed Veo-3.1 at 0.379) while all score about 0.8 on VBench; momentum is near zero for every model. Four VLMs asked to judge PASS/FAIL on real and synthetic anti-physics clips average 0.50 to 0.66 agreement. Curator qualifications: Curator located the excerpted evidence in the work copies (automated prefix match) and spot-checked claim-critical figures by search. The internal rounding/value inconsistencies are kept. Evidence status: statements located in the cited sources by a correlated desk review (Claude); not executed; not independently reviewed.
Structured fields
{
  "citations": [
    {
      "url": "https://arxiv.org/abs/2609.04200v1",
      "note": "abstract; Table 4"
    },
    {
      "url": "https://principiabench.github.io/",
      "note": "abstract; Table 4"
    },
    {
      "url": "https://github.com/PrincipiaBench/Principia",
      "note": "abstract; Table 4"
    }
  ]
}

Evidence claims

  • [sourced_fact] Principia (arXiv v1, 3 Sep 2026; project page says accepted to NeurIPS 2026) filmed paired objects under matched conditions (two blocks of different mass on identical inclines, hollow versus solid cylinders, pendulums of different lengths) and asks whether generated videos keep the pair's invariant (equal acceleration, equal period ratio, distance ordering). A unit-free score S = 1 - |F1 - F2| / (|F1| + |F2|) is computed in pixel space after SAM3 tracking. Six generators score between 0.148 and 0.419 Overall (Wan2.2-14B best, ahead of closed Veo-3.1 at 0.379) while all score about 0.8 on VBench; momentum is near zero for every model. Four VLMs asked to judge PASS/FAIL on real and synthetic anti-physics clips average 0.50 to 0.66 agreement. · locator: abstract; Table 4 source

Rights

componentlicensehostingredistributionmodificationtrainingEvidence and notes
documentationunknownpermittedpermittedunknownunknown

Applies to this original summary only, not to the cited sources.

Linked records

Machine-readable: JSON · Markdown · revisions · changes