Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1495 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Coding | Coverage (test) | 21 | Feb 26, 2026 | ||
| Classification | Coverage (test) | 21 | Feb 26, 2026 | ||
| Summarization | Coverage dataset (test) | 21 | Feb 26, 2026 | ||
| Knowledge Editing | Counterfact Full (test) | 21 | Feb 26, 2026 | ||
| Knowledge Editing | ZsRE (evaluation) | 21 | Feb 26, 2026 | ||
| Multi-hop Question Answering | IIRC | 21 | Jul 3, 2026 | ||
| Image Classification | CDDB-Hard (test) | 21 | Feb 26, 2026 | ||
| Image Captioning | MSCOCO (random sampling of 500 instances) | 21 | Feb 26, 2026 | ||
| Question Answering | HotpotQA December 2018 Wikipedia dump (test) | 21 | Jun 26, 2026 | ||
| Question Answering | NQ, TriviaQA, and WebQ (test) | 21 | Feb 26, 2026 | ||
| Long-context understanding | LongBench v1 (test) | 21 | Feb 26, 2026 | ||
| Instruction Following | Alpaca (test) | 21 | Apr 30, 2026 | ||
| Novel View Synthesis | NeRF Synthetic 800 x 800 (test) | 21 | Feb 26, 2026 | ||
| Novel View Synthesis | Real Forward-facing 640 x 960 (test) | 21 | Feb 26, 2026 | ||
| Video Understanding Reasoning | MLVU | 21 | Apr 10, 2026 | ||
| Reading Comprehension | RACE | 21 | Jun 4, 2026 | ||
| Long-form generation | LongBench Write-en | 21 | Jun 9, 2026 | ||
| Image Denoising | CBSD68 sigma=50 | 21 | Feb 26, 2026 | ||
| Image Denoising | CBSD68 sigma=25 | 21 | Feb 26, 2026 | ||
| Image Denoising | CBSD68 sigma=15 | 21 | Feb 26, 2026 | ||
| Jailbreak Defense | Aggregate Benchmarks | 21 | Feb 26, 2026 | ||
| Jailbreak Defense | GPTFuzzer | 21 | Feb 26, 2026 | ||
| Jailbreak Defense | ReNeLLM & DeepInception Average | 21 | Feb 26, 2026 | ||
| Fingerprint Verification | WizardMath-7B V1.0 | 21 | Feb 26, 2026 | ||
| Video Question Answering | DeVE-QA (test) | 21 | Feb 26, 2026 |