NVIDIA's Physis-Lang puts physics reasoning in captions, tops Physics-IQ with Cosmos 3
NVIDIAAI · x · 2026-09-30
NVIDIA, MIT, and Oxford released Physis-Lang, an open self-evolving framework that turns video captions into explicit physics representations.
- Captions decompose scenes into cause, governing physical law, and effect — explaining why a scene unfolds, not just what happens
- A physics-aware critic flags missing or unsupported claims; an agent refines shared captioning guidelines while the captioner stays fixed
- One representation connects data curation, fine-tuning, and inference: adding physics reasoning to prompts alone lifted Cosmos 3's PhyGenBench score by 5.62 points without retraining
- On the Physics-IQ Verified image-to-video leaderboard, Physis-Lang + Cosmos3 Super and Nano rank #1 (48.2) and #2 (43.3), beating the previous best of 42.7; Sora 2 sits at #20 with 26.5
Related event: NVIDIA's Physis-Lang Tops Physics-IQ Video Benchmark(4 posts)→
More from Multimodal
- New ChatGPT image failure mode: faces warp on handstand poses — flowersslop · 2026-11-15
- Fable 5.2 composes a piano piece 'Elsewhere, Then Here' and animates its own video — Antique_Cat4404 · 2026-11-15
- Creator claims no AI video platform makes money: Sora shut down, Grok Imagine layoffs — AIandDesign · 2026-11-15
- Use handheld camera feel instead of steadicam to add realism to AI video scenes — moplee1990 · 2026-11-15
- Coding over image models: pure-code art generation in Gemini and the app's nano banana bias — andrew_n_carr · 2026-11-15
- Midjourney style code 2659353360 yields striking ink-splash Marvel art — azed_ai · 2026-11-15