Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 2411 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Survival Analysis | FLC | 15 | May 22, 2026 | ||
| Survival Analysis | Rotterdam | 15 | May 22, 2026 | ||
| Long-term Memory Question Answering | ATM-Bench 1.0 (test) | 15 | May 6, 2026 | ||
| Visual Tracking | LasHeR | 15 | May 6, 2026 | ||
| 3D Anomaly Detection | Real3D-AD 1.0 (test) | 15 | May 6, 2026 | ||
| Traffic forecasting | England | 15 | May 6, 2026 | ||
| Neural PDE rollout | PDEBench Public metadata-hidden protocol (test) | 15 | May 6, 2026 | ||
| Edge Detection | BSDS500 | 15 | Jun 11, 2026 | ||
| Imitation Learning | Acrobot v1 (test) | 15 | May 6, 2026 | ||
| Imitation Learning | Cartpole v1 (test) | 15 | May 6, 2026 | ||
| Imitation Learning | Lunar Lander v2 (test) | 15 | May 6, 2026 | ||
| Node Classification | Coauthor-CS | 15 | May 6, 2026 | ||
| Law-level retrieval | Austrian-law benchmark (test) | 15 | May 6, 2026 | ||
| Law-level Retrieval | Austrian-law (test) | 15 | May 6, 2026 | ||
| Law-level retrieval | Austrian-law | 15 | May 6, 2026 | ||
| Jailbreaking | Multi-field Harmful Instruction Dataset 20 fields fused with 5 harmful instructions each | 15 | May 6, 2026 | ||
| Motion deblurring | RealBlur-J (test) | 15 | May 5, 2026 | ||
| Retrieval | TREC DL22 | 15 | May 4, 2026 | ||
| Discrete Branching Dynamics Reconstruction | Dyngen t=4 | 15 | May 4, 2026 | ||
| Discrete Branching Dynamics Reconstruction | Dyngen t=3 | 15 | May 4, 2026 | ||
| Discrete Branching Dynamics Reconstruction | Dyngen t=2 | 15 | May 4, 2026 | ||
| Discrete Branching Dynamics Reconstruction | Dyngen t=1 | 15 | May 4, 2026 | ||
| Semantic Segmentation | MultiSenGe 10m GSD (test) | 15 | May 4, 2026 | ||
| Semantic Segmentation | OSM 10m GSD (test) | 15 | May 4, 2026 | ||
| Semantic Segmentation | DW 10m GSD (test) | 15 | May 4, 2026 |