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 |
|---|---|---|---|---|---|
| Face Reconstruction | ORL face database (second subject) | 30 | Jun 5, 2026 | ||
| Deontic Reasoning | DeonticBench USCIS-AAO | 30 | Jun 4, 2026 | ||
| Deontic Reasoning | DeonticBench SARA-Bin | 30 | Jun 4, 2026 | ||
| Deontic Reasoning | DeonticBench Airline | 30 | Jun 4, 2026 | ||
| Deontic Reasoning | DeonticBench SARA-Num | 30 | Jun 4, 2026 | ||
| Jailbreak defense and Utility evaluation |
| Implicit risk dataset Management |
| 30 |
| Jun 4, 2026 |
| Jailbreak defense and Utility evaluation | Implicit risk dataset Finance | 30 | Jun 4, 2026 |
|---|
| Jailbreak defense and Utility evaluation | Implicit risk dataset Education | 30 | Jun 4, 2026 |
|---|
| Mathematical problem solving | GSM8K (test) | 30 | Jun 4, 2026 |
|---|
| Python program synthesis | MBPP (test) | 30 | Jun 4, 2026 |
|---|
| Cross-model transfer attack accuracy | CIFAR-10 (test) | 30 | Jun 4, 2026 |
|---|
| Survival Analysis | FLCHAIN | 30 | Jun 15, 2026 |
|---|
| Image Recognition | Asirra | 30 | Jun 4, 2026 |
|---|
| Face Recognition | KDEF | 30 | Jun 4, 2026 |
|---|
| Face Recognition | CASIA-Face V5 | 30 | Jun 4, 2026 |
|---|
| Scientific Reasoning | GPQA | 30 | Jun 30, 2026 |
|---|
| Preference Alignment | OpenRLHF Mixture | 30 | Jun 4, 2026 |
|---|
| Preference Alignment | Argilla | 30 | Jun 4, 2026 |
|---|
| 3D Bin Packing Problem | 3D-BPP ID-Small Discrete | 30 | Jun 4, 2026 |
|---|
| 3D Bin Packing Problem | 3D-BPP ID-Large Discrete | 30 | Jun 4, 2026 |
|---|
| Classification | labelme | 30 | Jul 7, 2026 |
|---|
| Multi-hop reasoning | MuSiQue Open Book (test) | 30 | Jun 4, 2026 |
|---|
| Random Field Reconstruction | Random Field Model Example 3 Eq. (29) (test) | 30 | Jun 4, 2026 |
|---|
| Language Model Downstream Evaluation | LM Evaluation Harness zero-shot and five-shot | 30 | Jun 2, 2026 |
|---|
| LLM Safety Evaluation | HARMAMP Benign instances | 30 | Jun 2, 2026 |
|---|