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 Synthesis | OCTMNIST | 48 | May 29, 2026 | ||
| Image Synthesis | CIFAR-10 | 48 | May 29, 2026 | ||
| Circle Packing | AlphaEvolve Circle Packing n=26 | 48 | May 29, 2026 | ||
| Axis Prediction Regression | 6-axis manipulator robot dataset Use-case 1 1 (test) | 48 | May 28, 2026 | ||
| Adversarial Robustness | CIFAR-10 |
| 48 |
| May 27, 2026 |
| Video Temporal Grounding | Charades-STA | 48 | May 28, 2026 |
|---|
| Medical Classification | ODIR (3-fold cross-validation) | 48 | May 26, 2026 |
|---|
| Medical Classification | Derm7pt | 48 | May 26, 2026 |
|---|
| Jailbreak Attack Evaluation | SafeBench 100 sampled harmful queries | 48 | May 26, 2026 |
|---|
| Image Classification | Fashion-MNIST (F-MNIST) | 48 | May 26, 2026 |
|---|
| Visual Prompt Injection Defense | ImgHijack Leak Context | 48 | May 26, 2026 |
|---|
| Multi-turn Tool-use | BFCL multi-turn V3 | 48 | Jun 18, 2026 |
|---|
| Multiclass Classification | Migraine clinical dataset six-class configuration (5-fold cross-validation) | 48 | May 25, 2026 |
|---|
| Speculative Decoding | GPQA Diamond | 48 | May 22, 2026 |
|---|
| Speculative Decoding | AIME 2024 | 48 | May 22, 2026 |
|---|
| Image-to-Text Adversarial Attack | Evaluation set | 48 | May 22, 2026 |
|---|
| Targeted Adversarial Attack | 1,000-pair Targeted Attack Evaluation Set closed-source standard MLLMs 1.0 | 48 | May 22, 2026 |
|---|
| Targeted Adversarial Attack | Evaluation set (test) | 48 | May 22, 2026 |
|---|
| Image Classification | ImageYellow ImageNet-1K subset (test) | 48 | May 21, 2026 |
|---|
| Image Classification | ImageMeow ImageNet-1K (test) | 48 | May 21, 2026 |
|---|
| Image Classification | ImageFruit ImageNet-1K (test) | 48 | May 21, 2026 |
|---|
| Knowledge Reasoning | GPQA Diamond | 48 | Jun 16, 2026 |
|---|
| Node classification | Cora, CiteSeer, PubMed, WikiCS, and Amazon Photo Downstream Average | 48 | May 20, 2026 |
|---|
| Node Classification | Physics (test) | 48 | May 19, 2026 |
|---|
| Node Classification | Cora (test) | 48 | May 19, 2026 |
|---|