Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1069 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| NL-to-FOL Syntax Correctness | FOLIO (test) | 26 | Feb 26, 2026 | ||
| Conversational Recommendation | REDDIT V2 (test) | 26 | Feb 26, 2026 | ||
| Multi-label abnormality classification | RAD-ChestCT (test) | 26 | Jun 9, 2026 | ||
| Severity Degree Diagnosis | ACL Sprain Diagnosis | 26 | Feb 26, 2026 | ||
| Severity Degree Diagnosis | Cartilage Thickness Loss (test) | 26 | Feb 26, 2026 | ||
| Retrieval | BelebeleRetrieval | 26 | May 19, 2026 | ||
| Image Quality Assessment | LIVEC | 26 | May 19, 2026 | ||
| Sequential Decision Making | ALFWorld (test) | 26 | Mar 13, 2026 | ||
| Time-series interval forecasting | Temperature | 26 | Jun 26, 2026 | ||
| Feature Selection | PIONeeR post-VIF | 26 | Feb 26, 2026 | ||
| Clustering | dermatology | 26 | May 12, 2026 | ||
| Gaussian deblurring | FFHQ (val) | 26 | May 27, 2026 | ||
| Audio Reconstruction | LJSpeech | 26 | Feb 26, 2026 | ||
| Language Identification | VoxLingua33 | 26 | Mar 10, 2026 | ||
| Speaker Identification | LibriSpeech MF | 26 | Mar 10, 2026 | ||
| Speech Spoofing Detection | In-the-Wild (ITW) (eval) | 26 | Jun 9, 2026 | ||
| Continual Learning | LIBERO-Long (test) | 26 | Jun 4, 2026 | ||
| Trigger Detection | ProAct-75 1.0 | 26 | Feb 26, 2026 | ||
| Object Navigation | HM3D (val) | 26 | Jun 15, 2026 | ||
| Reinforcement Learning | Ant v4 | 26 | Jun 26, 2026 | ||
| Constrained Optimization | BBOB-constrained 10D Problems | 26 | Feb 26, 2026 | ||
| Commonsense Question Answering | CSQA | 26 | Jul 9, 2026 | ||
| Multi-image visual perception | BLINK | 26 | Apr 15, 2026 | ||
| QoS Prediction (Throughput) | WS-DREAM Density=2.5% (test) | 26 | Feb 26, 2026 | ||
| Safety Moderation | Harmbench | 26 | Apr 7, 2026 |