Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 1999 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Traveling Salesman Problem | TSP500 Uniform distribution, scale ≤ 1,000 | 17 | May 26, 2026 | ||
| Dynamic Link Prediction | UCI-Message (Live-update) | 17 | May 26, 2026 | ||
| Fine-grained Understanding | FG-OVD | 17 | May 26, 2026 | ||
| Node Classification | Ele-Photo (test) | 17 | May 26, 2026 | ||
| Subject-driven personalization | DreamBench | 17 | May 26, 2026 | ||
| CTR Prediction | Average across all four datasets (Criteo, Avazu, KDD12, Industrial) | 17 | May 26, 2026 | ||
| Proactive Action Prediction | Home Loan | 17 | May 26, 2026 | ||
| Passage Retrieval | Amharic Passage Retrieval Dataset V2 (test) | 17 | May 26, 2026 | ||
| Web Research | BrowseComp | 17 | May 26, 2026 | ||
| Long-horizon agent performance | BrowseComp-ZH | 17 | May 26, 2026 | ||
| Textual understanding | Arena-Hard | 17 | May 26, 2026 | ||
| Textual understanding | AlpacaEval 2 | 17 | May 26, 2026 | ||
| Regression | BindingDB | 17 | May 26, 2026 | ||
| Classification | DrugBank | 17 | May 26, 2026 | ||
| Anomaly detection | SWaT | 17 | Jun 2, 2026 | ||
| Multivariate Time Series Anomaly Detection | PSM | 17 | May 29, 2026 | ||
| Question Answering | NQ PIA (test) | 17 | May 25, 2026 | ||
| Question Answering | HotpotQA Poison (test) | 17 | May 25, 2026 | ||
| Lesion Grounding and Segmentation | AbdomenAtlas 3.0 | 17 | May 22, 2026 | ||
| Proactive Task-Oriented Dialogue | FoodDeliverService Courier Regional Bonus | 17 | May 22, 2026 | ||
| Proactive Task-Oriented Dialogue | FoodDeliverService Merchant Promotion | 17 | May 22, 2026 | ||
| Video Anomaly Detection | ShanghaiTech Human-Related SHT-HR | 17 | May 22, 2026 | ||
| Video Temporal Grounding | Charades TimeLens-Bench (test) | 17 | May 22, 2026 | ||
| Video Temporal Grounding | QVHighlights TimeLens-Bench (test) | 17 | May 22, 2026 | ||
| Multimodal Embodied Reasoning | BEAR 1.0 (test) | 17 | May 22, 2026 |