Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 2356 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Human Preference Evaluation | User Study (Conversational Recommendation) (Human Evaluation) | 15 | Jun 24, 2026 | ||
| 3D Tumor Analysis | AbdomenAtlas 3.0 (test) | 15 | Jun 24, 2026 | ||
| Knowledge Retention | ImageNet | 15 | Jun 24, 2026 | ||
| Task Unlearning | 8 vision tasks | 15 | Jun 24, 2026 | ||
| Linear Mode Connectivity | ImageNet-1K (val) | 15 | Jun 23, 2026 | ||
| Fine-grained 9-way ill-posedness classification | CLAMBER 9-way | 15 | Jun 23, 2026 | ||
| Agentic Search | Overall | 15 | Jun 23, 2026 | ||
| Agentic Search | BrowseComp-Plus | 15 | Jun 23, 2026 | ||
| Agentic Search | BrowseComp | 15 | Jun 23, 2026 | ||
| Betti number prediction | Mantra 2-M^0 | 15 | Jun 23, 2026 | ||
| Image Classification | SVHN | 15 | Jun 30, 2026 | ||
| rPPG Estimation | UBFC -> MMPD | 15 | Jun 23, 2026 | ||
| rPPG Estimation | PURE -> UBFC | 15 | Jun 23, 2026 | ||
| rPPG Estimation | UBFC -> PURE | 15 | Jun 23, 2026 | ||
| Visual Understanding | MME Perception | 15 | Jun 23, 2026 | ||
| Agentic Model Routing | In-Distribution n=2,919 (test) | 15 | Jun 23, 2026 | ||
| Model Routing for Coding Tasks | Agent-as-a-Router In-distribution (test) | 15 | Jun 23, 2026 | ||
| Harm Recognition | Violence imagery dataset | 15 | Jun 23, 2026 | ||
| Harm Recognition | COCO Crime Scene slice | 15 | Jun 23, 2026 | ||
| Harm Recognition | Open Images Weapon slice | 15 | Jun 23, 2026 | ||
| Harm Recognition | Facebook Hateful Memes | 15 | Jun 23, 2026 | ||
| Harm Recognition | DeepGHS NSFW | 15 | Jun 23, 2026 | ||
| Multimodal Safety | Multimodal Safety Suite Avg | 15 | Jun 23, 2026 | ||
| Multimodal Safety | BeaverTails V | 15 | Jun 23, 2026 | ||
| Multimodal Safety | MM-Safety | 15 | Jun 23, 2026 |