Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1991 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Lifelong Unlearning | MLUBench Task C v1.0 (test) | 17 | Jun 12, 2026 | ||
| No-Reference Image Quality Assessment | Weighted average (SPAQ, AGIQA, LIVEW, KADID, PIPAL, TID2013, CSIQ) | 17 | Jun 12, 2026 | ||
| No-Reference Image Quality Assessment | AG-IQA | 17 | Jun 12, 2026 | ||
| Multi-Hop Question Answering | Bamb | 17 | Jun 12, 2026 | ||
| Multi-Hop Question Answering | 2wiki | 17 | Jun 12, 2026 | ||
| General Question Answering | HpQA | 17 | Jun 12, 2026 | ||
| General Question Answering | PopQA | 17 | Jun 12, 2026 | ||
| General Question Answering | TriviaQA | 17 | Jun 12, 2026 | ||
| General Question Answering | NQ | 17 | Jun 12, 2026 | ||
| General Question Answering | Natural Questions (NQ) v1.0 (test) | 17 | Jun 12, 2026 | ||
| Closed-loop Autonomous Driving | Bench2Drive official (test) | 17 | Jun 11, 2026 | ||
| Spatial Reasoning | ERQA | 17 | Jun 30, 2026 | ||
| Subject-preserving video generation | OpenS2V-Eval Human-Domain | 17 | Jun 11, 2026 | ||
| Text-to-image | CVTG-2K | 17 | Jun 11, 2026 | ||
| Forecasting | Lorenz-96 N=40 (test) | 17 | Jun 11, 2026 | ||
| Multi-hop Question Answering | Multi-hop QA Benchmarks (2WikiMultiHopQA, Bamboo, HotpotQA, MuSiQue, Natural Questions, PopQA, TriviaQA) | 17 | Jun 11, 2026 | ||
| Speaker Extraction | VoxCeleb2 mix Clean Visual | 17 | Jun 11, 2026 | ||
| Object Navigation | HM3D v1 (test) | 17 | Jun 11, 2026 | ||
| Code Generation | MBPP | 17 | Jun 11, 2026 | ||
| Code Generation | HumanEval | 17 | Jun 11, 2026 | ||
| Combinatorial Optimization | TSP n=150 Generalization | 17 | Jun 11, 2026 | ||
| Combinatorial Optimization | TSP n=125 Generalization | 17 | Jun 11, 2026 | ||
| Combinatorial Optimization | TSP n=100 (train) | 17 | Jun 11, 2026 | ||
| Image Classification | ImageNet sub-tasks S7-S9 (50 classes) | 17 | Jun 11, 2026 | ||
| Image Classification | ImageNet sub-tasks S4-S6 (25 classes) | 17 | Jun 11, 2026 |