Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 2344 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Text Quality Meta-evaluation | SummEval & Topical-Chat Combined (Overall) | 16 | Feb 26, 2026 | ||
| Text Quality Meta-evaluation | Topical-Chat (Local) | 16 | Feb 26, 2026 | ||
| Text Quality Meta-evaluation | SummEval (Local) | 16 | Feb 26, 2026 | ||
| Dialogue Response Generation | Topical-Chat Global | 16 | Feb 26, 2026 | ||
| Text Summarization | SummEval Global | 16 | Feb 26, 2026 | ||
| Phase Retrieval | FFHQ 256x256 (test) | 16 | Jun 5, 2026 | ||
| Anomaly Detection | BUSI | 16 | Feb 26, 2026 | ||
| Anomaly Localization | BUSZS | 16 | Feb 26, 2026 | ||
| Anomaly Localization | BUSI | 16 | Feb 26, 2026 | ||
| Natural Language Inference | XNLI French (test) | 16 | Feb 26, 2026 | ||
| Prompt Injection Defense | Prompt Injection Attacks (test) | 16 | Feb 26, 2026 | ||
| Object Composition | COCO (val) | 16 | Feb 26, 2026 | ||
| Domain Conversion | ImageNet to ImageNet-R (val) | 16 | Feb 26, 2026 | ||
| Graph property prediction | PCQM4M-LSC (train) | 16 | Feb 26, 2026 | ||
| Tool-based multi-turn dialogue | ToolWOZ (test) | 16 | Feb 26, 2026 | ||
| Knowledge Editing | RIPPLEEDITS single-instance | 16 | Feb 26, 2026 | ||
| Knowledge Editing | filtered dataset original (test) | 16 | Feb 26, 2026 | ||
| Knowledge Editing | in-prompt | 16 | Feb 26, 2026 | ||
| Knowledge Editing | replaced | 16 | Feb 26, 2026 | ||
| Semantic Segmentation | MaSS 13K 1,500 (test) | 16 | Feb 26, 2026 | ||
| Semantic Segmentation | MaSS13K 500 (val) | 16 | Feb 26, 2026 | ||
| Dependency Parsing | English (en) (test) | 16 | Feb 26, 2026 | ||
| Long-context understanding | LongBench v0.2 | 16 | Feb 26, 2026 | ||
| Few-shot Object Detection | Pascal VOC (Novel Split 1) | 16 | Feb 26, 2026 | ||
| Multitask Language Understanding | MMLU-pro | 16 | Feb 26, 2026 |