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 |
|---|---|---|---|---|---|
| Personalized Prediction | MovieLens (test) | 32 | Feb 26, 2026 | ||
| Personalized Scholarly Refinement | LaMP-5 | 32 | Feb 26, 2026 | ||
| Scientific Reasoning | MMLU-Pro | 32 | Apr 24, 2026 | ||
| Scientific Reasoning | GPQA Diamond | 32 | Feb 26, 2026 | ||
| Opinion Summarization Metric Evaluation | OpinSummEval | 32 | Feb 26, 2026 | ||
| Column Type Annotation |
| PublicBI to GitTables |
| 32 |
| Feb 26, 2026 |
| Reward Modeling | RM-Bench Hard | 32 | Jun 4, 2026 |
|---|
| Language Modeling | AdaptEval (test) | 32 | Feb 26, 2026 |
|---|
| Claim Verification | MR2 | 32 | Feb 26, 2026 |
|---|
| Claim Verification | Mocheg | 32 | Feb 26, 2026 |
|---|
| End-to-end throughput | Cloud Inference Workloads | 32 | Feb 26, 2026 |
|---|
| N-Queens problem | N-Queens | 32 | Feb 26, 2026 |
|---|
| Item crafting | TextCraft (test) | 32 | Feb 26, 2026 |
|---|
| Deepfake Detection | P2V | 32 | Feb 26, 2026 |
|---|
| Deepfake Detection | JDD 1.0 (test) | 32 | Feb 26, 2026 |
|---|
| Deepfake detection | SYN (test) | 32 | Feb 26, 2026 |
|---|
| Multi-hop Question Answering | HotpotQA | 32 | Feb 26, 2026 |
|---|
| Trajectory-level safety evaluation | R-judge (test) | 32 | Jun 2, 2026 |
|---|
| Task Efficient Routing | Software scenario | 32 | Feb 26, 2026 |
|---|
| Coding | HumanEval | 32 | Mar 18, 2026 |
|---|
| Code Verification | LiveCodeBench | 32 | Feb 26, 2026 |
|---|
| Code Verification | MBPP+ | 32 | Feb 26, 2026 |
|---|
| Code Verification | HumanEval+ | 32 | Feb 26, 2026 |
|---|
| Reasoning | Reasoning Evaluation Suite AIME 2024, GSM8k, MATH 500, GPQA | 32 | Feb 26, 2026 |
|---|
| General Reasoning | MMStar | 32 | Feb 26, 2026 |
|---|