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 |
|---|---|---|---|---|---|
| Image Classification | ImageNet | 40 | Jun 11, 2026 | ||
| Faithfulness Evaluation | ImageNet-1K | 40 | May 19, 2026 | ||
| Textual Explanation | ImageNet-1K misclassified cases | 40 | May 19, 2026 | ||
| Untargeted Adversarial Attack | ImageNet | 40 | May 1, 2026 | ||
| Vision Model Representation Alignment | ImageNet 20% held-out split of 20,000 samples (val) |
| 40 |
| Apr 23, 2026 |
| OOD Detection | ImageNet-1k (ID) vs SUN (OOD) | 40 | May 26, 2026 |
|---|
| Attribution Localization | ImageNet-1K ILSVRC2012 (val) | 40 | Mar 17, 2026 |
|---|
| Attribution Faithfulness | ImageNet-1K ILSVRC2012 (val) | 40 | Mar 17, 2026 |
|---|
| Image Classification | ImageNet | 40 | Apr 6, 2026 |
|---|
| VAE metric correlation with diffusion generation quality | ImageNet (val) | 40 | Mar 9, 2026 |
|---|
| Zero-Shot Posterior Sampling | IMAGENET 256x256 (val) | 40 | Feb 26, 2026 |
|---|
| Untargeted white-box adversarial attack | ImageNet | 40 | Feb 26, 2026 |
|---|
| Image Classification | ImageNet-1k 1.0 (val) | 40 | Feb 26, 2026 |
|---|
| Interpretation Error | ImageNet (test) | 40 | Feb 26, 2026 |
|---|
| Interpretation error | ImageNet (val) | 40 | Feb 26, 2026 |
|---|
| Out-of-Distribution Detection | ImageNet-1k V2 (test) | 40 | Feb 26, 2026 |
|---|
| Class-Incremental Learning | Split ImageNet-R 10 incremental tasks | 40 | May 13, 2026 |
|---|
| Clustering | ImageNet-Dogs (test) | 40 | Mar 26, 2026 |
|---|
| Image Classification | ImageNet 1K (train val) | 40 | May 26, 2026 |
|---|
| Backdoor Defense | ImageNet-12 (test) | 40 | Feb 26, 2026 |
|---|
| Image Classification | ImageNet-C 1.0 (val) | 40 | Feb 27, 2026 |
|---|
| Image Classification | ImageNet v1 (val) | 40 | Feb 26, 2026 |
|---|
| Image Classification | ImageNet-1K linear evaluation 1.0 | 40 | Feb 26, 2026 |
|---|
| Incremental Learning | ImageNet-1000 | 40 | Feb 26, 2026 |
|---|
| Low-shot Image Classification | ImageNet 1k (val) | 40 | Feb 26, 2026 |
|---|