PhysVista: Benchmarking Physical Intelligence in VLMs via a Perception-Reasoning-Assessment Loop
PhysVista tests whether VLMs can perceive, reason about, and judge physical plausibility in real and generated video.
PhysVista benchmarks physical intelligence in vision-language models through a closed loop of physical state perception, physical dynamics reasoning, and physical plausibility assessment. It distinguishes event-level and scale-level reasoning and includes both real-world and AI-generated videos. Experiments across a diverse set of VLMs find substantial limitations in physical reasoning and plausibility assessment, highlighting a gap between visual recognition and genuine physical understanding.
- Closed loop of perception, dynamics reasoning, and plausibility assessment
- Separates event-level and scale-level physical reasoning
- Covers both real-world and AI-generated videos
- VLMs show a persistent gap in physical understanding
Full article186 words · extracted from huggingface.co · click to collapse
Vision-Language Models (VLMs) have shown strong multimodal reasoning capabilities, yet whether they truly capture the physical consistency underlying real-world dynamics remains unclear. Existing benchmark paradigms often suffer from fragmented evaluation, focusing on isolated cognitive stages while overlooking the inherent synergy between perception, reasoning, and physical judgment. The lack of a holistic perspective limits the ability to diagnose whether VLMs can reliably evaluate the physical authenticity of emerging generative models. To address these issues, we introduce PhysVista, a benchmark designed to evaluate physical intelligence in VLMs through a closed cognitive loop framework inspired by the human seeing-reasoning-assessment process. PhysVista restores this loop by jointly evaluating physical state perception, physical dynamics reasoning, and physical plausibility assessment. It further distinguishes event-level reasoning and scale-level reasoning to enable fine-grained analysis of physical understanding. In addition, PhysVista incorporates both real-world and AI-generated videos, allowing evaluation across diverse domains and emerging generative scenarios. Extensive experiments across a diverse set of VLMs reveal substantial limitations in physical reasoning and plausibility assessment, highlighting a persistent gap between visual recognition and genuine physical understanding, and pointing toward more principled designs for physically grounded multimodal intelligence.
Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/papers/2610.00559