Benchmarks

Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.

309,131 indexed

Task	Dataset	Domain	Results	Last Update
Long-context language understanding	LongBench-e	llm chat instructionnlp	93	May 27, 2026
Novel View Synthesis	ScanNet++	3d vision geometry	93	Jul 1, 2026
Anomaly Detection	BraTS	tabular time series	93	Jun 15, 2026
Multivariate Time Series Anomaly Detection	SMAP	tabular time series	93	Jun 23, 2026
Image Classification	Average of 11 datasets (ImageNet, Caltech101, OxfordPets, StanfordCars, Flowers102, Food101, FGVCAircraft, SUN397, DTD, EuroSAT, UCF101) Base-to-Novel Generalization	computer vision