Paper Detail
Bingxuan Li, Yining Hong, Cheng Qian, Hyeonjeong Ha, Jiateng Liu, Zhenhailong Wang, Yue Guo, Yunzhu Li, Heng Ji
Physical interactions follow a long-tailed distribution: a set of common and regular interactions dominates human experience and visual data, while a broad spectrum of rare and irregular interactions remains underrepresented. Although recent visual world models, including image and video generation models, achieve impressive realism on existing benchmarks, they primarily focus on simulating common physical interactions. This raises a central question: Do current visual world models internalize and generalize physical principles? In this work, we introduce Tailor-Bench, a benchmark that challenges world models to simulate irregular physical interactions. To enable systematic evaluation, we design three scenario modes that progressively challenge model reasoning: Regular scenarios reflect common tool-task pairs, Unconventional scenarios replace conventional tools with attribute-compatible substitutes to test affordance generalization, and Impossible scenarios introduce attribute-violating tools to probe constraint awareness. Additionally, we design two complementary settings under a unified evaluation protocol: predictive generation requires inferring outcomes without guidance, while descriptive generation specifies the target outcome for faithful realization. Our experimental results reveal a clear long-tail gap in physical world modeling: performance degrades from Regular to Unconventional and Impossible scenarios, indicating limited generalization beyond common interactions. Failure analysis further shows that models rely on superficial visual patterns: image models fail to realize correct state changes, while video models further suffer from temporal inconsistencies.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@misc{li2026trimming,
title = {Trimming the Long-Tail of Visual World Modeling Evaluation},
author = {Bingxuan Li and Yining Hong and Cheng Qian and Hyeonjeong Ha and Jiateng Liu and Zhenhailong Wang and Yue Guo and Yunzhu Li and Heng Ji},
year = {2026},
abstract = {Physical interactions follow a long-tailed distribution: a set of common and regular interactions dominates human experience and visual data, while a broad spectrum of rare and irregular interactions remains underrepresented. Although recent visual world models, including image and video generation models, achieve impressive realism on existing benchmarks, they primarily focus on simulating common physical interactions. This raises a central question: Do current visual world models internalize a},
url = {https://huggingface.co/papers/2606.24256},
keywords = {visual world models, physical interactions, long-tailed distribution, image generation, video generation, world model evaluation, scenario modes, regular scenarios, unconventional scenarios, impossible scenarios, predictive generation, descriptive generation, physical principle generalization, affordance generalization, constraint awareness, temporal consistency, code available, huggingface daily},
eprint = {2606.24256},
archiveprefix = {arXiv},
}
{}