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 |
|---|---|---|---|---|---|
| Visual Jailbreaking Attack | HADES | 32 | Mar 4, 2026 | ||
| Visual Jailbreaking Attack | AdvBench | 32 | Mar 4, 2026 | ||
| Commonsense Reasoning | Commonsense Reasoning (PIQA, WinoG., HellaS., BoolQ, SIQA, OBQA) (test) | 32 | Feb 27, 2026 | ||
| Image Classification | StanfordCars, DTD, EuroSAT, GTSRB, RESISC45, SUN397, SVHN | 32 | Jul 1, 2026 | ||
| Text Classification | 20News (test) |
| 32 |
| Mar 11, 2026 |
| Time Series Forecasting | ETTm1 v1 (test) | 32 | Apr 28, 2026 |
|---|
| Online Inference | ShareGPT | 32 | May 27, 2026 |
|---|
| Multilingual Understanding | CMMLU | 32 | May 27, 2026 |
|---|
| Membership Inference Attack | GitHub | 32 | Mar 23, 2026 |
|---|
| Multivariate Time-series Forecasting | Traffic | 32 | Feb 26, 2026 |
|---|
| Multivariate Time-series Forecasting | Electricity | 32 | Feb 26, 2026 |
|---|
| Contributor Attribution | CIFAR-20 | 32 | Feb 26, 2026 |
|---|
| Mathematical Reasoning | Minerva | 32 | Mar 4, 2026 |
|---|
| Instruction Following | AlpacaEval 2 (val) | 32 | Jun 4, 2026 |
|---|
| Multi-class Classification | EOL (test) | 32 | Feb 26, 2026 |
|---|
| Multi-class Classification | IF (test) | 32 | Feb 26, 2026 |
|---|
| Multi-class Classification | JV (test) | 32 | Feb 26, 2026 |
|---|
| Binary Classification | Employee (test) | 32 | Feb 26, 2026 |
|---|
| Binary Classification | KC1 (test) | 32 | Feb 26, 2026 |
|---|
| Conversational Question Answering | COQA zero-shot (test) | 32 | Feb 26, 2026 |
|---|
| Mathematics Question Answering | GSM8K zero-shot (test) | 32 | Feb 26, 2026 |
|---|
| Causal Discovery | Synthetic n=1000, |E|=2000, sample size=1000 | 32 | Feb 26, 2026 |
|---|
| Regression | Housing | 32 | May 12, 2026 |
|---|
| Zero-shot Question Answering and Reasoning | LM-Eval-Harness Suite (PIQA, HellaSwag, LAMBADA, ARC-e, ARC-c, SciQ, Race, MMLU) zero-shot | 32 | Feb 26, 2026 |
|---|
| Recommendation | Baby | 32 | Apr 7, 2026 |
|---|