Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 282 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Representational accuracy | 14 main-study subjects 1.0 (test) | 60 | May 29, 2026 | ||
| Code correctness prediction | MultiPL-E Java | 60 | May 28, 2026 | ||
| Predicting code correctness | LiveCodeBench Python | 60 | May 28, 2026 | ||
| Code Correctness Prediction | MultiPL-E Java | 60 | May 28, 2026 | ||
| Code correctness prediction | LiveCodeBench Python | 60 | May 28, 2026 | ||
| Code Correctness Prediction | MultiPL-E Java | 60 | May 28, 2026 | ||
| Code Correctness Prediction | LiveCodeBench Python | 60 | May 28, 2026 | ||
| Code Generation | MBPP Code | 60 | May 28, 2026 | ||
| Mathematical Reasoning | AIME24 Math | 60 | May 28, 2026 | ||
| Targeted Adversarial Attack | PhyDB | 60 | May 28, 2026 | ||
| Meningioma grade classification | MEN (test) | 60 | May 26, 2026 | ||
| Image Classification | ImageNet-1k | 60 | Jul 7, 2026 | ||
| Question Answering under PIA attack | Bio | 60 | May 22, 2026 | ||
| Question Answering | TQA | 60 | May 22, 2026 | ||
| Body Composition Segmentation | SAROS held-out (test) | 60 | May 21, 2026 | ||
| Robotic Task Completion | Kinetix | 60 | May 20, 2026 | ||
| Continual Unlearning | ImageNet-1K | 60 | May 19, 2026 | ||
| Mathematical reasoning | AIME 2024 | 60 | Jun 15, 2026 | ||
| Feature Importance Ranking | Simulated binary classification 1.0 (train test) | 60 | May 15, 2026 | ||
| Explainable AI Attribution | ImageNet random subset of 1000 images | 60 | May 15, 2026 | ||
| Interactive Decision Making | Textcraft | 60 | Jun 4, 2026 | ||
| Chinese character recognition | 1000 (test) | 60 | May 12, 2026 | ||
| Safety Alignment Breaking Prevention | StrongREJECT | 60 | May 12, 2026 | ||
| Safety Alignment Breaking Prevention | HarmBench | 60 | May 12, 2026 | ||
| Harmful Knowledge Evaluation | WMDP evil | 60 | May 12, 2026 |