Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1496 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Image-Text Compositionality Evaluation | SugarCrepe ++ (test) | 21 | Mar 27, 2026 | ||
| Multimodal reasoning | VisuLogic | 21 | Jun 9, 2026 | ||
| Video Super-Resolution | REDS | 21 | Feb 26, 2026 | ||
| Zero-shot Evaluation | StableEval 27 evals | 21 | Feb 26, 2026 | ||
| Classification | Scene-Centric Datasets | 21 | Feb 26, 2026 | ||
| Classification | Object-Centric datasets | 21 | Feb 26, 2026 | ||
| Super-Resolution | HGGT (test-100) | 21 | Feb 26, 2026 | ||
| Stereo Matching | Middlebury full resolution | 21 | Apr 23, 2026 | ||
| Citation-augmented Question Answering | GT, PK 1.0 (test) | 21 | Feb 27, 2026 | ||
| Long-term forecasting | ETTh1 T=96 | 21 | Jun 1, 2026 | ||
| Safety Alignment | XSTest | 21 | May 12, 2026 | ||
| Video Reconstruction | Panda-70M | 21 | Apr 21, 2026 | ||
| Person Re-identification | LTCC Clothing Change (test) | 21 | Feb 26, 2026 | ||
| Machine Translation | En-Ru document-level | 21 | Mar 27, 2026 | ||
| Time Series Forecasting | ETTm1 zero-shot transfer from ETTh1 | 21 | Apr 15, 2026 | ||
| Few-Shot Action Recognition | HMDB | 21 | Feb 26, 2026 | ||
| Backdoor Attack | STL10 | 21 | Mar 31, 2026 | ||
| Perception Evaluation | MME Perception | 21 | Mar 25, 2026 | ||
| Implicit Image Representation | DIV2K (train test) | 21 | Feb 26, 2026 | ||
| Cardiac ultrasound segmentation | CardiacUDA (test) | 21 | Feb 26, 2026 | ||
| Anomaly Classification | MTD (test) | 21 | Feb 26, 2026 | ||
| Concept Erasure | CUB (test) | 21 | Feb 26, 2026 | ||
| Concept Erasure | Oxford Flowers (test) | 21 | Feb 26, 2026 | ||
| Concept Erasure | Stanford Dogs (test) | 21 | Feb 26, 2026 | ||
| Language Compositionality | SugarCrepe (test) | 21 | Feb 26, 2026 |