Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1784 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Image Classification | CIFAR-10 | 18 | Jul 8, 2026 | ||
| Nodule & Tumor Segmentation | TN3K | 18 | Jul 8, 2026 | ||
| Explanatory sentence ranking | Legal Case-Law Overall | 18 | Jul 8, 2026 | ||
| Novel View Synthesis | MegaScenes fixed (test) | 18 | Jul 7, 2026 | ||
| Novel View Synthesis | Phototourism (test) | 18 | Jul 7, 2026 | ||
| Formal Language Modeling | Formal Language Tasks (val) | 18 | Jul 7, 2026 | ||
| Information Extraction and Aggregation | Web DOM Nanobrowser based synthesized element format | 18 | Jul 7, 2026 | ||
| Information Extraction and Aggregation | JSON production APIs and synthetic reviews | 18 | Jul 7, 2026 | ||
| Multi-task evaluation summary | Aggregate MATH-500, AIME, MBPP+, HE+, MMLU | 18 | Jul 7, 2026 | ||
| Language Understanding | MMLU | 18 | Jul 7, 2026 | ||
| General Evaluation | UltraFeedback Aggregate | 18 | Jul 7, 2026 | ||
| General Language Understanding | MMLU | 18 | Jul 7, 2026 | ||
| Mathematical Reasoning | MATH 500 | 18 | Jul 7, 2026 | ||
| Efficiency and Optimization Evaluation | Multi-Objective Efficiency Evaluation | 18 | Jul 7, 2026 | ||
| Fairness Evaluation | Co-occurrence tests (CooC) | 18 | Jul 7, 2026 | ||
| Question Answering | AI2 Reasoning Challenge Easy | 18 | Jul 7, 2026 | ||
| Trajectory Generation | San Francisco | 18 | Jul 7, 2026 | ||
| Trajectory Generation | Beijing | 18 | Jul 7, 2026 | ||
| Terminal interaction agent task | Terminal-Bench (dev) | 18 | Jul 7, 2026 | ||
| Image Classification | MNIST (test) | 18 | Jul 7, 2026 | ||
| 3D Object Detection | OmniHD-Scenes (test) | 18 | Jul 7, 2026 | ||
| Language Modeling | C4 en (train) | 18 | Jul 7, 2026 | ||
| Multi-domain Evaluation | MMLU, GSM8K, HEval, BBH, MedQA | 18 | Jul 7, 2026 | ||
| Taste Prediction | Unified music-taste corpus (test) | 18 | Jul 7, 2026 | ||
| Weakly-Supervised Semantic Segmentation | BraTS Edema 2020 (val) | 18 | Jul 7, 2026 |