MMLU-by-task - High School Statistics — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 1257 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 2 70B Base | 63.43 |
| 2 | falcon-180B | 62.04 |
| 3 | LLaMA-65B | 60.65 |
| 4 | Llama 2 70B Chat | 59.72 |
| 5 | Mistral-7B-v0.1 | 57.41 |
| 6 | StableBeluga2 | 56.94 |
| 7 | OpenHermes-13B | 50 |
| 8 | LLaMA-30B | 48.61 |
| 9 | Llama 2 70B Chat GPTQ | 48.61 |
| 10 | Llama 2 70B Chat (HF) | 48.15 |
| 11 | CodeLlama-13B-hf | 48.15 |
| 12 | CodeLlama-7B-hf | 47.22 |
| 13 | bloom-1b7 | 47.22 |
| 14 | CodeLlama-7B-Python-hf | 46.3 |
| 15 | vicuna-7B-v1.3 | 45.83 |
Interactive version: theaggregate.ai/benchmark?slug=mmlu-by-task-high-school-statistics · How the rankings work · Data refreshed daily, snapshot 2026-07-22.