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 | OBQA | 36 | Feb 27, 2026 | ||
| Commonsense Reasoning | HellaSwag | 36 | Feb 27, 2026 | ||
| Boolean Question Answering | BoolQ | 36 | Feb 27, 2026 | ||
| Fine-grained Image Classification | DTD (novel classes) | 36 | Feb 27, 2026 | ||
| Fine-grained Image Classification | FGVCAircraft (novel classes) | 36 | Feb 27, 2026 | ||
| Image Classification |
| Food101 novel classes |
| 36 |
| May 12, 2026 |
| Watermark Extraction | Geometric Distortions (test) | 36 | Feb 26, 2026 |
|---|
| Time Series Reasoning | TRQA | 36 | Jun 12, 2026 |
|---|
| Time Series Reasoning | TSQA | 36 | Jun 12, 2026 |
|---|
| Time Series Reasoning | ETI | 36 | Jun 12, 2026 |
|---|
| Accurate Retrieval | Accurate Retrieval (AR) suite | 36 | Feb 26, 2026 |
|---|
| Test-Time Learning | Test-Time Learning (TTL) suite | 36 | Feb 26, 2026 |
|---|
| General Task | MMStar | 36 | Jun 2, 2026 |
|---|
| Graph Edit Distance Approximation | AIDS | 36 | Feb 26, 2026 |
|---|
| Language Modeling | AG News (val) | 36 | May 26, 2026 |
|---|
| Mathematical Reasoning | GSM-Symbolic Origin 100 | 36 | Feb 26, 2026 |
|---|
| Mathematical Reasoning | GSM-Symbolic Vary Both | 36 | Feb 26, 2026 |
|---|
| Mathematical Reasoning | GSM-Symbolic Vary Name | 36 | Feb 26, 2026 |
|---|
| Mathematical Reasoning | GSM-Symbolic Vary Num. | 36 | Feb 26, 2026 |
|---|
| Sequential Recommendation | Office | 36 | May 28, 2026 |
|---|
| Mutual Information Estimation | Out-of-Meta-Distribution (OoMD) (test) | 36 | Feb 26, 2026 |
|---|
| Mutual Information Estimation | In-Meta-Distribution (IMD) (test) | 36 | Feb 26, 2026 |
|---|
| Downstream NLP Evaluation | Downstream NLP Tasks (Lambada, ARC, WinoGrande, PIQA, HellaSwag, SciQ, RACE) LM Evaluation Harness (test) | 36 | Feb 26, 2026 |
|---|
| Cross-modal registration | Replica 62 | 36 | Feb 27, 2026 |
|---|
| Biomedical Multi-hop Question Answering | CondMedQA | 36 | Feb 26, 2026 |
|---|