Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1502 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Image Classification | FMNIST | 21 | Feb 26, 2026 | ||
| Mathematical Reasoning | GSM8K | 21 | Feb 26, 2026 | ||
| Motion Forecasting | nuScenes official leaderboard (test) | 21 | Feb 26, 2026 | ||
| Multi-image reasoning | BLINK (val) | 21 | Feb 26, 2026 | ||
| Multi-image reasoning | QBench2 (val) | 21 | Feb 26, 2026 | ||
| Information Retrieval | ClueWeb 500K | 21 | Feb 26, 2026 | ||
| Information Retrieval | Gov 500K | 21 | Feb 26, 2026 | ||
| LaMP-1 Personalization | LaMP-1 (val) | 21 | Jun 2, 2026 | ||
| Image Classification | CIFAR10 (test) | 21 | Feb 26, 2026 | ||
| Long-context Question Answering | HotpotQA | 21 | Feb 26, 2026 | ||
| Long Dependency Question Answering | LooGLE | 21 | Apr 14, 2026 | ||
| Saliency Prediction | SalECI E-Commercial | 21 | Jun 4, 2026 | ||
| Image Classification | ImageNet-1K 1.0 (val) | 21 | Feb 26, 2026 | ||
| LLM Inference Latency | 256 tokens generation (test) | 21 | Feb 26, 2026 | ||
| Medical Image Segmentation | Covid-19 | 21 | Feb 26, 2026 | ||
| Question Answering | StrategyQA | 21 | Feb 26, 2026 | ||
| Multivariate long-term forecasting | ETTm2 (Avg) | 21 | Apr 21, 2026 | ||
| Object Hallucination Evaluation | POPE MSCOCO (val) | 21 | Feb 26, 2026 | ||
| Unconditional Image Generation | ImageNet 256x256 (train) | 21 | Feb 26, 2026 | ||
| Prostate Segmentation | Prostate MRI 6 institutions (test) | 21 | Feb 26, 2026 | ||
| Natural Language Inference | GNLI Human (test) | 21 | Feb 26, 2026 | ||
| Natural Language Inference | WANLI (test) | 21 | Feb 26, 2026 | ||
| Natural Language Inference | ANLI (val) | 21 | Feb 26, 2026 | ||
| Reasoning Segmentation | MUSE (val) | 21 | Feb 26, 2026 | ||
| Semantic Segmentation | ADE20k overlapped setting (100-10 (6 steps)) | 21 | Feb 26, 2026 |