Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1573 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Self-Preference Bias Analysis | AlpacaEval | 20 | Apr 28, 2026 | ||
| GUI Action Prediction | GUI-Odyssey | 20 | May 12, 2026 | ||
| Code Reasoning | CRUX official (test) | 20 | Apr 28, 2026 | ||
| Correspondence-free point cloud registration | Stanford 3D scanning repository | 20 | Apr 27, 2026 | ||
| Clustering | S3 N(200) | 20 | Apr 27, 2026 | ||
| Clustering | S2 N(150) | 20 | Apr 27, 2026 | ||
| Clustering | S1 N(150) | 20 | Apr 27, 2026 | ||
| Healthcare provider taxonomy code classification | Healthcare Taxonomy | 20 | Apr 27, 2026 | ||
| Organization-level SIC code classification | SIC Code | 20 | Apr 27, 2026 | ||
| Mispronunciation Detection | L2-ARCTIC (test) | 20 | Apr 27, 2026 | ||
| Daily Irradiance Forecasting | Solaris Real Experiments Location 4 | 20 | Apr 27, 2026 | ||
| Attribution | ImageNet (val) | 20 | May 20, 2026 | ||
| Attribution | COCO (val) | 20 | Apr 27, 2026 | ||
| LLM Attack Effectiveness Evaluation | LLM Attack Effectiveness Dataset | 20 | Apr 24, 2026 | ||
| Protein structure generation | Foldseek AFDB clusters DFS (test) | 20 | Apr 24, 2026 | ||
| Surgical Video Question Answering | EndoVis18-VQA Out-of-template | 20 | Jun 17, 2026 | ||
| Surgical Video Question Answering | EndoVis-VQA In-template 18 | 20 | Jun 17, 2026 | ||
| Scientific Poster Generation | Paper2Poster | 20 | Jul 7, 2026 | ||
| Optic Disc Segmentation | Drishti-GS (external val) | 20 | Apr 24, 2026 | ||
| Optic Disc Segmentation | IDRID | 20 | Apr 24, 2026 | ||
| Optic Disc Segmentation | GoDARTS | 20 | Apr 24, 2026 | ||
| LLM evaluation correctness | PPE Correctness track | 20 | Apr 24, 2026 | ||
| Jailbreaking | Harmful prompt (test) | 20 | Jun 16, 2026 | ||
| Sequential Recommendation | Ebnerd | 20 | Apr 24, 2026 | ||
| Sequential Recommendation | MicroVideo | 20 | Apr 24, 2026 |