SCORE Robustness (Accuracy): leaderboard
NVIDIA SCORE evaluates robustness on AGIEval, MATH, and MMLU-Pro under benchmark perturbations, with this view averaging task accuracy.
Metric: Average Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.1 70B Instruct | 67.02 | #561 |
| 2 | Mistral Large 2 (Jul) | 66.97 | #484 |
| 3 | Llama 3 70B Instruct | 58.67 | #634 |
| 4 | Yi 1.5 34B Chat | 54.16 | #800 |
| 5 | Llama 3.1 8B Instruct | 48.41 | #1035 |
| 6 | Mistral Nemo Instruct (2407) | 44.72 | #936 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=score-robustness-accuracy · How It Works · Data refreshed daily, snapshot 2026-10-20.