Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 2432 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| Image Captioning Evaluation | LongCap-Arena (test) | 15 | Apr 7, 2026 | ||
| Image Captioning Evaluation | MMHE (test) | 15 | Apr 7, 2026 | ||
| Sequential Recommendation | Beauty | 15 | Apr 7, 2026 | ||
| Behavioral Cloning | P2P-Dyn. | 15 | Apr 7, 2026 | ||
| Active Multi-target Tracking | Active Multi-Target Tracking Simulation (5 episodes) | 15 | Apr 7, 2026 | ||
| Query-document retrieval | MAIR (20% labeled) | 15 | Apr 7, 2026 | ||
| Retrieval | MAIR (test) | 15 | Apr 7, 2026 | ||
| Retrieval | MAIR 10% labeled query-document pairs | 15 | Apr 7, 2026 | ||
| Information Retrieval | MAIR 5% labeled query-document pairs (test) | 15 | Apr 7, 2026 | ||
| Retrieval | MAIR 40% labeled query-document pairs | 15 | Apr 7, 2026 | ||
| Common-sense Reasoning | WinoGrande S In-Distribution Llama-3.1-8B (train test) | 15 | Apr 7, 2026 | ||
| Compliance Verification | IPv6 standards compliance 500 samples (test) | 15 | Apr 7, 2026 | ||
| Question Answering | SciQAG | 15 | Apr 7, 2026 | ||
| Question Answering | RepliQA 1,000 QA pairs (test) | 15 | Apr 7, 2026 | ||
| Question Answering | SciQAG 1,000 sampled questions (test) | 15 | Apr 7, 2026 | ||
| C-index alignment between cosine similarity and GPT-4 assigned relevance scores | Clinical Disease Feature Relevance Suite (T1D, Epilepsy, PH, Asthma, CD, UC) | 15 | Apr 7, 2026 | ||
| Mathematical Reasoning | MATH500 | 15 | Apr 6, 2026 | ||
| Mathematical Reasoning | AIME 2025 | 15 | Apr 6, 2026 | ||
| Reasoning | GPQA Diamond | 15 | Apr 6, 2026 | ||
| Person Re-Identification | SYSU-CMCC Infrared to Visible | 15 | Apr 6, 2026 | ||
| Person Re-Identification | SYSU-CMCC Visible to Infrared | 15 | Apr 6, 2026 | ||
| Image Segmentation | UPENN-GBM | 15 | May 15, 2026 | ||
| Potential gradient estimation | Reference model | 15 | Apr 6, 2026 | ||
| Medical Visual Question Answering | Pooled Medical VQA Open-ended | 15 | Apr 6, 2026 | ||
| Revenue Uplift Modeling | Production Dataset | 15 | Apr 6, 2026 |