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 |
|---|---|---|---|---|---|
| Scene Graph Classification | Action Genome (test) | 55 | Apr 23, 2026 | ||
| Medical Visual Question Answering | PathVQA (test) | 55 | May 26, 2026 | ||
| Automatic Speech Recognition | GigaSpeech (test) | 55 | Jul 2, 2026 | ||
| FoV regression | Cars3D (all) | 55 | Feb 26, 2026 | ||
| Graph Anomaly Detection | T-Finance |
| 55 |
| Jun 9, 2026 |
| Graph Understanding | GraphSQA Overall 1.0 (test) | 55 | Feb 26, 2026 |
|---|
| Graph Understanding | GraphSQA Micro-level Tasks 1.0 (test) | 55 | Feb 26, 2026 |
|---|
| Graph Understanding | GraphSQA Macro-level Tasks 1.0 (test) | 55 | Feb 26, 2026 |
|---|
| Single Image Super-Resolution | Set5 (test) | 55 | Feb 26, 2026 |
|---|
| Visible-Infrared Image Fusion | MSRS (test) | 55 | Jun 16, 2026 |
|---|
| Jailbreaking | HARMBENCH 159 standard behaviors (test) | 55 | Jun 23, 2026 |
|---|
| Video Understanding | EgoSchema (test) | 55 | Mar 20, 2026 |
|---|
| Sequential model editing | ZsRE | 55 | Apr 7, 2026 |
|---|
| Image Classification | ImageNet Natural Distribution Shifts suite (ImageNet, ImageNet-A, ImageNet-V2, ImageNet-R, ImageNet-Sketch) (test) | 55 | Jun 30, 2026 |
|---|
| Dynamic Facial Expression Recognition | DFEW | 55 | Jun 9, 2026 |
|---|
| Multi-modal Understanding | MMBench EN | 55 | Mar 10, 2026 |
|---|
| Shape correspondence estimation | TOPKIDS | 55 | Jun 26, 2026 |
|---|
| Jailbreak Defense | AutoDAN | 55 | Apr 20, 2026 |
|---|
| Video Semantic Segmentation | VSPW | 55 | Apr 21, 2026 |
|---|
| Semantic Textual Similarity | BIOSSES | 55 | Apr 20, 2026 |
|---|
| Summarization Evaluation | SummEval | 55 | Jun 26, 2026 |
|---|
| Change Detection | WHU-CD | 55 | Feb 26, 2026 |
|---|
| 3D Semantic Segmentation | 3D-OVS | 55 | Jul 3, 2026 |
|---|
| Semantic Segmentation | ACDC (Night) | 55 | Jun 2, 2026 |
|---|
| BEV Semantic Segmentation | nuScenes (val) | 55 | May 26, 2026 |
|---|