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 |
|---|---|---|---|---|---|
| Ordinal Regression | Helpsteer | 48 | May 12, 2026 | ||
| Object Detection | Multi-center Internal real (test) | 48 | May 12, 2026 | ||
| Reasoning | LSAT | 48 | May 12, 2026 | ||
| Selective Classification | Unseen Cls. shift | 48 | May 12, 2026 | ||
| Question Answering | RULER | 48 | Jun 4, 2026 | ||
| Code-Specific Instruction Tuning Evaluation |
| Magicoder Evaluation Suite |
| 48 |
| May 8, 2026 |
| Nonlinear equality-constrained optimization | Randomly generated nonlinear equality-constrained problem instances threshold 10^-3 (out of five random instances per condition level) | 48 | May 8, 2026 |
|---|
| Science | MMLU-Pro | 48 | Jul 2, 2026 |
|---|
| Code Generation | MBPP+ | 48 | Jun 18, 2026 |
|---|
| Aggregate Reconstruction | NHANES 8 templates (holdout) | 48 | May 6, 2026 |
|---|
| Multi-objective optimization | DTLZ6 | 48 | May 6, 2026 |
|---|
| Multi-objective optimization | DTLZ4 | 48 | May 6, 2026 |
|---|
| Multi-objective optimization | DTLZ3 | 48 | May 6, 2026 |
|---|
| Multi-objective optimization | DTLZ2 | 48 | May 6, 2026 |
|---|
| Multi-objective optimization | DTLZ1 | 48 | May 6, 2026 |
|---|
| Anomaly Detection | TELCO (test) | 48 | May 4, 2026 |
|---|
| Voronoi regret minimization | Unit Sphere (S^{d-1}) synthetic setup | 48 | May 4, 2026 |
|---|
| Multivariate Time Series Forecasting | Traffic | 48 | May 1, 2026 |
|---|
| Multivariate Time Series Forecasting | Electricity | 48 | May 1, 2026 |
|---|
| Multivariate Time Series Forecasting | Traffic | 48 | May 1, 2026 |
|---|
| Image Classification | MNIST | 48 | May 1, 2026 |
|---|
| Query Performance Prediction | TREC DL'19 and DL'20 (2-fold train-test) | 48 | May 1, 2026 |
|---|
| Anomaly Detection | TELCO | 48 | May 4, 2026 |
|---|
| Perceptual Audio Quality Assessment | Vibrating Objects | 48 | Apr 30, 2026 |
|---|
| Negative Sentiment Backdoor Detection | Gemma 2 9B | 48 | Apr 28, 2026 |
|---|