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 |
|---|---|---|---|---|---|
| Spatio-temporal video grounding | HCSTVG v1 (test) | 42 | Apr 15, 2026 | ||
| Video Action Recognition | UCF101 avg over all splits | 42 | Feb 26, 2026 | ||
| Location and orientation estimation | VIGOR Same-Area | 42 | Jun 11, 2026 | ||
| Collaborative Filtering | Yelp 2018 | 42 | Feb 26, 2026 | ||
| Multiple Choice Question Answering |
| BIG-bench HHH Eval |
| 42 |
| Jun 2, 2026 |
| Few-shot Segmentation | Pascal 5^i (test) | 42 | Feb 26, 2026 |
|---|
| Text-to-Image Generation | MS-COCO 2014 | 42 | Feb 26, 2026 |
|---|
| Image Classification | ImageNet RobustBench (val) | 42 | May 27, 2026 |
|---|
| Class Unlearning | CIFAR-10 (test) | 42 | May 18, 2026 |
|---|
| Image Classification | STL-10 40 labels | 42 | Apr 24, 2026 |
|---|
| Sound Event Tagging | FSD50K (test) | 42 | Jun 16, 2026 |
|---|
| No-Reference Video Quality Assessment | KoNViD-1k | 42 | Feb 26, 2026 |
|---|
| 3D Classification | ScanObjectNN PB-T50-RS official | 42 | Feb 26, 2026 |
|---|
| Visual Question Answering | AI2D | 42 | Mar 24, 2026 |
|---|
| Visual Question Answering | ST-VQA | 42 | Apr 21, 2026 |
|---|
| Image Denoising | CBSD68 sigma=50 (test) | 42 | Feb 26, 2026 |
|---|
| Mathematical Reasoning | GSM8k | 42 | Feb 26, 2026 |
|---|
| Single Object Tracking | LaSOT-ext | 42 | Mar 10, 2026 |
|---|
| Image Classification | CIFAR-100 Noise Ratio 50% | 42 | Feb 26, 2026 |
|---|
| Image Classification | CIFAR-100 Noise Ratio 20% | 42 | Feb 26, 2026 |
|---|
| Image Classification | CIFAR-10 Noise Ratio 50% | 42 | Feb 26, 2026 |
|---|
| Image Classification | CIFAR-10 Noise Ratio 20% | 42 | Feb 26, 2026 |
|---|
| Semantic Segmentation | Pascal Context 459 classes (val) | 42 | Feb 26, 2026 |
|---|
| Referring Video Object Segmentation | Ref-YouTube-VOS 2019 (val) | 42 | Apr 21, 2026 |
|---|
| Argument Identification and Classification | ERE-EN (dev) | 42 | Feb 26, 2026 |
|---|