Benchmarks

Task NameDataset NameSOTA ResultTrendResults

Runtime analysis

Synthetic dataset c=2

0.001CPU Runtime (s)

Multi-view Clustering

Synthetic Dataset Setup

1Rand Index

Novel View Synthesis

Synthetic Dataset (test)

58.38PSNR

Multi-camera Pose Optimization

Synthetic Dataset

0.25Runtime (s)

Counterfactual Prediction

Synthetic Dataset

0.538RMSEy

Counterfactual Prediction

Synthetic Dataset (test)

0.161PEHE

Counterfactual Prediction

Synthetic Dataset

0ATE Error

Local Lipschitz constant estimation

16-dimensional synthetic dataset (test)

628.91Estimated Lipschitz Constant

Attribute Inference

Synthetic Dataset

71.24Attribute Inference Accuracy

Future Link Prediction

Synthetic Dataset w/ OOD

70.12AUC

Multi-view Clustering

Synthetic Dataset Setup 2

1Rand Index

Operator Learning

Synthetic Dataset

0.01MAE

Multi-label content safety classification

Synthetic Dataset

0.768F1 Score

Calibration Testing

Synthetic dataset (test)

0.01Calibration Threshold

Guided Flow Matching

Synthetic Dataset 10k samples 2-dimensional Mixture of Gaussians

94.6Posterior Coverage

Censored-engression

Synthetic dataset Regime (D)

6.25Oracle Score

Change Point Detection

Synthetic Dataset 2 (test)

96.5F1 Score

Multiple Rotation Averaging

Synthetic dataset (test)

0.23Mean Angular Error

Revenue Uplift Modeling

Synthetic Dataset

0.316AUUC

Set Containment

Synthetic dataset (test)

100Accuracy

National-level electricity forecasting

Synthetic Dataset Σ = I (test)

1.75MAPE

Signal recovery

Unbalanced synthetic dataset (n=(50, 70, 120), N=20, K=3)

10.4SNR

Forecasting

Synthetic dataset 10 systems, 20 input steps, 1 step horizon (test)

0.241RMSE

Semantic Segmentation

Synthetic dataset (val)

0.735mIoU

Object Detection

Synthetic dataset (val)

66.9Precision