PAI-Bench: A Comprehensive Benchmark For Physical AI

arXiv 2025 37.8 benchmark

TLDR

Introduces PAI-Bench, a benchmark evaluating perception and prediction in Physical AI using 2,808 real-world cases across video tasks.

Reasoning

The paper's strength lies in its comprehensive benchmark with real-world data and task-aligned metrics for physical plausibility. However, it only evaluates existing models without proposing new methods, and its scope is limited to video tasks.

Read-first score

Read-first score 37.8, 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,metric

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

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 251.

Keyword Scores

world dynamics prediction

world model

video world model

world simulator

generative world model

interactive world model

model-based reinforcement learning world model

Resources

Paper