Rapidly craft compelling rankingusing prompts
OpenTAI Rank is exploring the most advanced language and visual models, offering a comprehensive, and neutral evaluation reference for the industry and research community.
Learn how it works
Rankings模型名称 | 论文 | 模型下载量 | 对抗安全性 | 排名 | |
|---|---|---|---|---|---|
| 数据集A | 数据集B | ||||
| GPT-4 | Robust Principles: Architectural DesignPrinciples for Adversarially Robust... | 112092 | 76.5% | 76.5% | 1 |
| Llama 3 | Better Diffusion Models Further lmprove Adversarial Training | 112092 | 76.5% | 76.5% | 2 |
| Claude 3 | MixedNUTS: Training-Free Accuracy-RobustnessBalance via Nonlinearly Mixed... | 112092 | 76.5% | 76.5% | 3 |
| Ernie-4.0 | Improving the Accuracy-Robustness Trade-off ofClassifiers via Adaptive Smoothing | 112092 | 76.5% | 76.5% | 4 |
| GPT-3.5 turbo | Better Diffusion Models Further lmprove Adversarial Training | 112092 | 76.5% | 76.5% | 5 |
| GPT-3.5 | MixedNUTS: Training-Free Accuracy-RobustnessBalance via Nonlinearly Mixed... | 112092 | 76.5% | 76.5% | 6 |
| xAI-Grok 2 | Better Diffusion Models Further lmprove Adversarial Training | 112092 | 76.5% | 76.5% | 7 |
Load more