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
Class-Incremental Learning	Class-Incremental Learning Sequences Easy	computer visiontabular time series	36	Mar 5, 2026
Class-Incremental Learning	Class-Incremental Learning Sequences Hard	computer visiontabular time series	36	Mar 5, 2026
Mathematical Reasoning	MinervaMath	llm chat instructionnlp	36	Apr 20, 2026
Scientific Code Generation	SciCode (test)	code developer ai	36	Mar 4, 2026
Speaker Impersonation Attack	VoxCeleb1 and CNCeleb subsampled (test-enrollment)	speech audioadversarial safety security