Apple-π benchmark asks whether video models reason about physical laws or just mimic motion
liuziwei7 · x · 2026-07-21
Apple-π benchmark tests whether video models really reason about physics
The post highlights Apple-π, described as the first law-grounded benchmark for video models.
It contains 400 videos and 10 mechanics tasks, and evaluates models through a Perception → Formulation → Deduction pipeline. The central question is whether video models actually reason about physical laws, or merely generate motion that looks plausible.
Related event: Apple-π Benchmark Tests Video Models' Physics Reasoning(3 posts)→
More from Models
- Fern 0.1 debuts: a model built to audit training data and debug agent traces for AI builders — bneyshabur · 2026-10-26
- Fern 0.1 posts 73.4 on SWE-Bench Verified, far behind frontier models on agentic benchmarks — bneyshabur · 2026-10-26
- Fern's "context is intelligence" thesis: prediction keeps improving out to 10.2M tokens — bneyshabur · 2026-10-26
- Mirendil launches Fern, a from-scratch pretrained model with up to 10M-token context — bneyshabur · 2026-10-26
- Users spot a dedicated translation UI appearing in Claude — BLCNYY · 2026-10-26
- GrokBot user: flight booking and shopping all blocked, agent can't do anything useful — lucasmeijer · 2026-10-26