Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 2631 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| End-to-end unlearning | TOFU 10% | 14 | May 27, 2026 | ||
| Node classification | BITalpha | 14 | May 27, 2026 | ||
| Node classification | TAX | 14 | May 27, 2026 | ||
| Node classification | DBLP | 14 | May 27, 2026 | ||
| Hate speech classification | MHS small models | 14 | May 27, 2026 | ||
| Hate Speech Classification | MHS | 14 | May 27, 2026 | ||
| Scientific Question Answering | GPQA | 14 | May 27, 2026 | ||
| Mathematical Reasoning | AMO decontaminated | 14 | May 27, 2026 | ||
| Mathematical Reasoning | AIME25 decontaminated | 14 | May 27, 2026 | ||
| Mathematical Reasoning | AIME decontaminated 24 | 14 | May 27, 2026 | ||
| Mathematical Reasoning | AMC23 decontaminated | 14 | May 27, 2026 | ||
| Mathematical Reasoning | HLE decontaminated | 14 | May 27, 2026 | ||
| Mathematical Reasoning | Olympiad decontaminated | 14 | May 27, 2026 | ||
| Mathematical Reasoning | Minerva decontaminated | 14 | May 27, 2026 | ||
| Mathematical Reasoning | Math500 decontaminated | 14 | May 27, 2026 | ||
| Unlearning | SD Preserved Concepts Set 2.1 | 14 | May 27, 2026 | ||
| Unlearning | SD Van Gogh Erased Set 2.1 | 14 | May 27, 2026 | ||
| Agent behavioral safety | InjecAgent | 14 | May 27, 2026 | ||
| Agent behavioral safety | AgentDojo | 14 | May 27, 2026 | ||
| Agent behavioral safety | AgentHarm | 14 | May 27, 2026 | ||
| TKR prediction | OAI target | 14 | May 27, 2026 | ||
| TKR prediction | OAI source | 14 | May 27, 2026 | ||
| Question Answering | NQ (val) | 14 | May 27, 2026 | ||
| Personalized Text Generation | Amazon Reviews Books, Movies_and_TV, and CDs_and_Vinyl 2023 | 14 | May 27, 2026 | ||
| Physics-informed quantitative forecasting | Helmholtz Staircase Equation (test) | 14 | May 27, 2026 |