A Physical Coherence Benchmark for Evaluating Video Generation Models via Optical Flow-guided Frame Prediction

arXiv 2025 49.3 benchmark, method

TLDR

Introduces PhyCoBench, a benchmark with 120 prompts across 7 physical law categories, plus PhyCoPredictor, to evaluate physical coherence of text-to-video generation models.

Reasoning

The paper presents a clearly motivated benchmark with manual and automated evaluations across multiple video generation models, and releases resources publicly. However, the visible abstract does not detail experimental protocols, statistical robustness, or limitations of the proposed automated predictor.

Read-first score

Read-first score 49.3, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 22.

Recency 8%

86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=benchmark,dataset,evaluation,experiment,result

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=dataset,github

Topical relevance 42%

31.4

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Field roles

FrontierMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 595.

Keyword Scores

world simulator

video world model

world model

world dynamics prediction

generative world model

interactive world model

model-based reinforcement learning world model

Resources

Paper