Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 995 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Composed Image Retrieval | GeneCIS | 27 | May 21, 2026 | ||
| Odometry | GRACO aerial-01 to aerial-08 | 27 | May 21, 2026 | ||
| Semantic Segmentation | Cityscapes | 27 | May 21, 2026 | ||
| ASR Error Correction | Common Voice Frisian (test) | 27 | May 20, 2026 | ||
| ASR Error Correction | Frisian Offline Data (test) | 27 | May 20, 2026 | ||
| First-error step detection | MoreHopQA | 27 | May 20, 2026 | ||
| Step-wise Confidence Attribution | Math | 27 | May 20, 2026 | ||
| Step-wise Confidence Attribution | MoreHopQA | 27 | May 20, 2026 | ||
| Step-wise Confidence Attribution | GSM8K | 27 | May 20, 2026 | ||
| Text-to-SQL | Spider 2.0 (test) | 27 | May 20, 2026 | ||
| Preference-aligned Retrieval-Augmented Generation | PrefEval | 27 | May 19, 2026 | ||
| Preference-aligned Retrieval-Augmented Generation | PrefELI5 | 27 | May 19, 2026 | ||
| Preference-aligned Retrieval-Augmented Generation | PrefRQ | 27 | May 19, 2026 | ||
| Preference-aligned Retrieval-Augmented Generation | PrefWiki | 27 | May 19, 2026 | ||
| Pansharpening | WV3 full-resolution | 27 | Jun 17, 2026 | ||
| Hate Speech Detection | HateXplain | 27 | May 19, 2026 | ||
| End-to-end GP fit+predict | Synthetic GP Data D=4 | 27 | May 19, 2026 | ||
| Language Modeling Evaluation | General Language Evaluation Suite AE, AC, SciQ, MMLU, MMLU-P, HS, OBQA, PIQA, RACE, WG, CSQA, AGI (test) | 27 | May 19, 2026 | ||
| Scientific Reasoning | GPQA-D | 27 | May 19, 2026 | ||
| Mathematical Reasoning | AIME 25 | 27 | May 19, 2026 | ||
| Mathematical Reasoning | AIME 24 | 27 | May 19, 2026 | ||
| Mathematical Reasoning | MATH 500 | 27 | May 19, 2026 | ||
| Polynomial multiplication | Fq[x]/<x^n + 1> | 27 | May 19, 2026 | ||
| Human Image Animation | RealisDance (val) | 27 | Jun 30, 2026 | ||
| General Reasoning | SuperGPQA | 27 | Jun 9, 2026 |