Research
Benchmarks
Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.
309,131 benchmarksPage 833 of 12,366
Benchmarks
| Task | Dataset | Domain | Trend | Results | Last update |
|---|---|---|---|---|---|
| LLM Safety Evaluation | HARMAMP (malicious instances) | 30 | Jun 2, 2026 | ||
| IMU-based state estimation | EuRoC MAV | 30 | Jun 2, 2026 | ||
| Graph Matching | Synthetic 2D point clouds | 30 | Jun 2, 2026 | ||
| Regression | California dataset | 30 | Jun 2, 2026 | ||
| Graph Alignment | Regular Random Graphs degree 10 | 30 | Jun 2, 2026 | ||
| Circuit Localization | IOI | 30 | Jun 2, 2026 | ||
| Knowledge Editing | RippleEdits POPULAR (full requested-edit set) | 30 | Jun 2, 2026 | ||
| Video Compression | JVET CTC Cb Component | 30 | Jun 2, 2026 | ||
| Video Compression | JVET CTC Cr component VTM-11.0 Anchor (All Classes) | 30 | Jun 2, 2026 | ||
| Novel View Synthesis | TAT Dataset (test) | 30 | Jun 2, 2026 | ||
| Fact verification | FEVER | 30 | Jun 2, 2026 | ||
| Action Recognition | HAC various domain shifts | 30 | Jun 2, 2026 | ||
| Meta-Reinforcement Learning | Hopper-Param (ID) | 30 | Jun 2, 2026 | ||
| Image Classification | TinyImageNet | 30 | Jun 2, 2026 | ||
| LLM Inference | Qwen3-8B 2k prompts Decode-heavy workload | 30 | Jun 2, 2026 | ||
| Agentic Search | Agentic Search Benchmarks Average | 30 | Jun 2, 2026 | ||
| Multi-hop Question Answering | MuSiQue | 30 | Jun 2, 2026 | ||
| Multi-hop Question Answering | 2WikiMultiHopQA | 30 | Jun 2, 2026 | ||
| Fréchet Regression | Simulated distributional outputs under linear relationships | 30 | Jun 2, 2026 | ||
| Concept Erasure | I2P | 30 | Jun 2, 2026 | ||
| Human Activity Recognition | Milan (mixed (pure + cross) windows) | 30 | Jun 2, 2026 | ||
| RLHF Backdoor Attack | Anthropic Helpful Harmless prompts (train test) | 30 | Jun 2, 2026 | ||
| MLLM Unlearning | MLLMU-Bench Retain Set 10% ratio | 30 | Jun 2, 2026 | ||
| MLLM Unlearning | MLLMU-Bench 10% ratio (test) | 30 | Jun 2, 2026 | ||
| MLLM Unlearning | MLLMU-Bench forget set, 10% ratio | 30 | Jun 2, 2026 |