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 |
|---|---|---|---|---|---|
| Multimodal Retrieval-Augmented Generation | MRAMG | 32 | Feb 26, 2026 | ||
| Interactive Agent Task | TRIP-Bench (Hard LIT) | 32 | Feb 26, 2026 | ||
| Mathematical Reasoning | Mathematical Reasoning Benchmarks (GSM8K, MATH, AMC23, Olympiad, Minerva) (test) | 32 | Feb 26, 2026 | ||
| Output Length Prediction | ForeLen RL | 32 | Feb 26, 2026 | ||
| Output Length Prediction | ForeLen Reasoning |
| 32 |
| Feb 26, 2026 |
| RTL generation | VerilogEval 156 cases (test) | 32 | Feb 26, 2026 |
|---|
| Novel View Synthesis | RealEstate-10K 2-view | 32 | Mar 23, 2026 |
|---|
| Uncertainty Estimation | NQ-open | 32 | Apr 21, 2026 |
|---|
| Goal-oriented Dialogue | Taxi | 32 | Apr 15, 2026 |
|---|
| Respiratory sound classification | ICBHI 2017 (official) | 32 | Feb 26, 2026 |
|---|
| Lifelong robotic skill learning | MimicGen Lifelong tasks | 32 | Feb 26, 2026 |
|---|
| Reasoning | Reasoning Tasks Average | 32 | Feb 26, 2026 |
|---|
| Question Answering | Overall NQ, TriviaQA, BioASQ, PopQA | 32 | Feb 26, 2026 |
|---|
| Red-teaming Safety Evaluation | Harmbench | 32 | Feb 26, 2026 |
|---|
| Semi-supervised Video Object Segmentation | EntitySeg (ES), PartImageNet-Universe (PU), LVIS (LV), and SA-V (SV) SAM2 VOS Evaluation Suite (test) | 32 | Feb 26, 2026 |
|---|
| Biomedical Evidence Synthesis | COCHRANEFOREST (test) | 32 | Feb 26, 2026 |
|---|
| Safety | Beavertails | 32 | Feb 26, 2026 |
|---|
| Fact-Checking | COVID-Fact | 32 | May 12, 2026 |
|---|
| Text Classification | LRA Text (test) | 32 | Feb 26, 2026 |
|---|
| Machine-Generated Text Detection | M4 | 32 | Feb 26, 2026 |
|---|
| Image Retrieval | SOP | 32 | Feb 26, 2026 |
|---|
| GUI Navigation | Multimodal-Mind2Web Cross-Domain | 32 | Apr 23, 2026 |
|---|
| GUI Navigation | Multimodal-Mind2Web Cross-Task | 32 | Apr 23, 2026 |
|---|
| UAV Tracking | UAVDT | 32 | Mar 31, 2026 |
|---|
| UAV Tracking | DTB70 | 32 | Mar 31, 2026 |
|---|