Datasets:
|
Download README.md from World-Representation-Lab/World-Embedding-Regression: direct link, hf CLI and curl.
- Browser
- Download file 5.39 kB
-
https://huggingface.co/datasets/World-Representation-Lab/World-Embedding-Regression/resolve/main/README.md
- Command line
-
hf download hf://datasets/World-Representation-Lab/World-Embedding-Regression/README.md
-
curl -L -o README.md https://huggingface.co/datasets/World-Representation-Lab/World-Embedding-Regression/resolve/main/README.md
5.39 kB
metadata
license: apache-2.0
task_categories:
- video-classification
tags:
- physics
- video-regression
- benchmark
- physical-laws
size_categories:
- 1K<n<10K
configs:
- config_name: pendulum
data_files:
- split: test
path: pendulum/test-*.parquet
- config_name: rigidCollision
data_files:
- split: test
path: rigidCollision/test-*.parquet
- config_name: sphereDrop3D
data_files:
- split: test
path: sphereDrop3D/test-*.parquet
- config_name: dye2d
data_files:
- split: test
path: dye2d/test-*.parquet
- config_name: flowpastcylinder2d
data_files:
- split: test
path: flowpastcylinder2d/test-*.parquet
- config_name: totalInternalReflection
data_files:
- split: test
path: totalInternalReflection/test-*.parquet
- config_name: doubleSlitInterference
data_files:
- split: test
path: doubleSlitInterference/test-*.parquet
- config_name: cylinderTorsion3D
data_files:
- split: test
path: cylinderTorsion3D/test-*.parquet
- config_name: longCuboidBending3D
data_files:
- split: test
path: longCuboidBending3D/test-*.parquet
- config_name: hyperelasticNeoHookeanCuboid3D
data_files:
- split: test
path: hyperelasticNeoHookeanCuboid3D/test-*.parquet
dataset_info:
- config_name: pendulum
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: rigidCollision
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: sphereDrop3D
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: dye2d
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: flowpastcylinder2d
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: totalInternalReflection
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: doubleSlitInterference
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: cylinderTorsion3D
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: longCuboidBending3D
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
- config_name: hyperelasticNeoHookeanCuboid3D
features:
- name: family
dtype: string
- name: id
dtype: string
- name: video
dtype: video
- name: regression_attribute
dtype: string
- name: regression_value
dtype: float64
splits:
- name: test
num_examples: 500
World Embedding Regression
World Embedding Regression is the physical-property regression dataset from the World Embedding Benchmark. It contains 5,000 simulation videos across 10 physics families, with 500 examples in each dataset configuration.
Usage
from datasets import load_dataset
dataset = load_dataset(
"World-Representation-Lab/World-Embedding-Regression",
"pendulum",
split="test",
)
Fields
family: physics family.id: unique example identifier.video: simulation video.regression_attribute: physical quantity to predict.regression_value: continuous target value.
The fixed test and scaling splits used by the benchmark are distributed with the evaluation code.
Citation
@misc{liu2026worldembeddingbenchmark,
title = {World Embedding Benchmark},
author = {Yiqi Liu and Ruifeng Yuan and Yang Wang and Long Li and Fengyu Cai and Hou Pong Chan and Jialin Yu and Hao Zhang and Chenghua Lin and Chenghao Xiao},
year = {2026},
eprint = {2610.03632},
archivePrefix = {arXiv},
primaryClass = {cs.CV},
url = {https://arxiv.org/abs/2610.03632}
}