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 |
|---|---|---|---|---|---|
| Question Answering | TriviaQA | 36 | Mar 18, 2026 | ||
| Image Classification | 14 tasks collection | 36 | Mar 18, 2026 | ||
| Image Classification | 8 tasks collection | 36 | Mar 18, 2026 | ||
| Binary Classification | MagicTel 20% | 36 | Mar 18, 2026 | ||
| Binary Classification | Haberman 10% | 36 | Mar 18, 2026 | ||
| Binary Classification | CreditCard 0.2% |
| 36 |
| Mar 18, 2026 |
| Fused Matmul and Sampling | Synthetic D=4096, V=151k | 36 | Mar 18, 2026 |
|---|
| Dialect Translation | Jeolla, Gyeongsang, and Jeju dialects averaged (test) | 36 | Mar 17, 2026 |
|---|
| Object Removal | RORD (val) | 36 | Jun 29, 2026 |
|---|
| Scientific Reasoning | SCIENTIFIC | 36 | Mar 17, 2026 |
|---|
| Mathematical Reasoning | MATH Easy | 36 | Mar 17, 2026 |
|---|
| Membership Inference | CLIP image-text (train) | 36 | Mar 17, 2026 |
|---|
| Video Anomaly Detection | XD-Violence | 36 | May 15, 2026 |
|---|
| Image Classification | ImageNet I-Woof (test) | 36 | May 6, 2026 |
|---|
| Physical Reasoning | Cut the Rope | 36 | Mar 17, 2026 |
|---|
| Physical Reasoning | Angry Birds | 36 | Mar 17, 2026 |
|---|
| Physical Reasoning | Pooltool | 36 | Mar 17, 2026 |
|---|
| Physical Reasoning | Kinetix | 36 | Mar 17, 2026 |
|---|
| Physical Reasoning | I-PHYRE DeepPHY (test) | 36 | Mar 17, 2026 |
|---|
| Physical Reasoning | PHYRE DeepPHY (test) | 36 | Mar 17, 2026 |
|---|
| No-Reference Image Quality Assessment | CLIVE | 36 | Jun 30, 2026 |
|---|
| Image Classification | Ultrahigh Carbon Steel Micrograph DataBase (UHCS) (IR=10, pl=0.1, α=0.1) 1.0 (test) | 36 | Mar 16, 2026 |
|---|
| Image Classification | Ultrahigh Carbon Steel Micrograph DataBase (UHCS) (IR=2, pl=0.1, α=0.1) 1.0 (test) | 36 | Mar 16, 2026 |
|---|
| Math Reasoning | AIME 24 | 36 | Mar 16, 2026 |
|---|
| Video Understanding | VideoMME w/o sub | 36 | Jun 11, 2026 |
|---|