Loading the SOTA2 catalog…
SOTA2 Research · benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
| Task | Dataset | Domain | Trend | Results | Last Update |
|---|---|---|---|---|---|
| Out-of-Distribution Detection | Places with ImageNet-1k OOD In-distribution (test) | 99 | Feb 26, 2026 | ||
| Video Summarization | TVsum | 99 | May 12, 2026 | ||
| Facial Expression Recognition | RAF-DB | 99 | Jun 23, 2026 | ||
| Cross-view geo-localization | CVACT (test) |
| 99 |
| Mar 30, 2026 |
| Image Denoising | DND (test) | 99 | Apr 21, 2026 |
|---|
| Super-Resolution | Manga109 4x | 99 | Mar 13, 2026 |
|---|
| Action Detection | AVA v2.2 (val) | 99 | Feb 26, 2026 |
|---|
| 5-way few-shot classification | CUB | 99 | Jun 5, 2026 |
|---|
| Action Recognition | HMDB51 (split 1) | 99 | Jun 23, 2026 |
|---|
| Dependency Parsing | Chinese Treebank (CTB) (test) | 99 | Feb 26, 2026 |
|---|
| Semantic Segmentation | NYUDv2 40-class (test) | 99 | Feb 26, 2026 |
|---|
| Text-to-Image Generation | T2I-CompBench++ | 99 | Jun 23, 2026 |
|---|
| Multi-person Pose Estimation | COCO 2017 (test-dev) | 99 | Feb 26, 2026 |
|---|
| Uncertainty Quantification | NQ | 98 | Jun 4, 2026 |
|---|
| Uncertainty Quantification | TQA | 98 | Jun 4, 2026 |
|---|
| Uncertainty Quantification | XSum | 98 | Jun 4, 2026 |
|---|
| Uncertainty Quantification | NQ 16 additional samples (test) | 98 | Jun 4, 2026 |
|---|
| Uncertainty Quantification | TQA 16 additional samples (test) | 98 | Jun 4, 2026 |
|---|
| Question Answering | NQ 8 additional samples | 98 | Jun 4, 2026 |
|---|
| Question Answering | TQA 8 additional samples | 98 | Jun 4, 2026 |
|---|
| Uncertainty Quantification | NQ 4 additional samples (test) | 98 | Jun 4, 2026 |
|---|
| Uncertainty Quantification | TQA 4 additional samples (test) | 98 | Jun 4, 2026 |
|---|
| Image Classification | CIFAR100 (test) | 98 | May 27, 2026 |
|---|
| Indirect Prompt Injection Defense | ICLR Academic Reviews 84 defensive payloads / 252 prompts 2026 | 98 | May 8, 2026 |
|---|
| Time Series Forecasting | Exchange | 98 | May 26, 2026 |
|---|