PhysicsLENS: Diagnosing Physical Property Blindness in Video Generation Models

arXiv:2610.01162v1 Announce Type: new Abstract: Reliable video world models could provide scalable predictive environments for robot learning, planning, and evaluation. However, generated robot videos can violate physical principles and complete tasks through physically implausible behavior, limiting their reliability for robot learning and planning. Current video-generation benchmarks exclude physics that are inherently hidden by visuals (e.g., weight, viscosity, friction). Due to this, video models are evaluated on the fidelity of physics, not the underlying accuracy of physics. We introduce PhysicsLENS, a dataset and benchmark for evaluating plausibility of physical properties grounded in robotics. PhysicsLENS uses matched scenario pairs that hold the same conditioning frame and task, while varying underlying physics in the scene description. Scenarios are curated from public robot video sources and annotated across seven physical domains: collision, gravity, momentum, friction, deformation, fluid, and causality. We evaluate across four video generation models, producing over 400 human-annotated labels. Results show that plausible-looking videos often ignore the stated property (34 of 47), and that stating the property lowers plausibility only slightly and not significantly.

Related

• Physics Question Scene Graph: Fine-grained Evaluation of Physical Plausibility in Text-to-Video Generation
• ICAT: Incident-Case-Grounded Adaptive Testing for Physical-Risk Prediction in Embodied World Models
• Each Judge Its Own Yardstick: Discovering Per-VLM Taxonomies for Physical Video Evaluation

Source: arXiv cs.CV | 2026-10-02