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 |
|---|---|---|---|---|---|
| Ordered Pixel-by-Pixel Classification | MNIST ordered pixels (test) | 42 | Feb 26, 2026 | ||
| Character-level Prediction | PTB (test) | 42 | Feb 26, 2026 | ||
| Point Cloud Quality Assessment | SJTU-PCQA (test) | 42 | Feb 26, 2026 | ||
| Semantic Segmentation | MFNet nighttime (test) | 42 | Feb 26, 2026 | ||
| Fine-grained classification | Pets (test) | 42 | Jun 8, 2026 | ||
| Reveal (test) |
| 42 |
| Feb 26, 2026 |
| Code Debugging | HumanEval | 42 | Feb 26, 2026 |
|---|
| Image Classification | CIFAR-10N (Random 1) | 42 | May 21, 2026 |
|---|
| Recommendation | Amazon-Movie IID (test) | 42 | Feb 26, 2026 |
|---|
| Dichotomous Image Segmentation | DIS5K TE (1-4) (test) | 42 | May 15, 2026 |
|---|
| Continual Image Classification | MiniImageNet Split | 42 | Mar 18, 2026 |
|---|
| Scene Classification | RESISC-45 (TR=20%) | 42 | May 12, 2026 |
|---|
| Object Hallucination Probing | GQA POPE Random | 42 | Apr 10, 2026 |
|---|
| Attention Operator Throughput | Llama2 7B (32 Q-heads/32 KV-heads/128 Head-dimension) | 42 | May 7, 2026 |
|---|
| Multitask Language Understanding | MMLU | 41 | Jul 7, 2026 |
|---|
| Math Problem Solving | AIME 2024 | 41 | Jun 26, 2026 |
|---|
| Class-conditioned generation | ImageNet 256 × 256 (train test) | 41 | Jun 25, 2026 |
|---|
| Instruction Following | IFBench | 41 | Jul 7, 2026 |
|---|
| Localization | CityU Construction | 41 | Jun 15, 2026 |
|---|
| Protein-Ligand Binding Affinity Prediction | PDBbind core set 2013 | 41 | Jun 15, 2026 |
|---|
| Multi-label recognition | VOC 2007 | 41 | Jun 12, 2026 |
|---|
| Multi-label recognition | COCO 2017 | 41 | Jun 12, 2026 |
|---|
| Mathematics | MATH-500 (test) | 41 | Jun 16, 2026 |
|---|
| Machine-Generated Text Detection | COLING MGT English 2025 | 41 | Jun 8, 2026 |
|---|
| Multi-Criteria Decision Making | Alternative set A augmented with A11 | 41 | Jun 8, 2026 |
|---|