MMLU-by-task - Sociology — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 1257 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 2 70B Base | 90.05 |
| 2 | StableBeluga2 | 89.55 |
| 3 | falcon-180B | 89.55 |
| 4 | Llama 2 70B Chat (HF) | 87.06 |
| 5 | Llama 2 70B Chat | 87.06 |
| 6 | Llama 2 70B Chat GPTQ | 84.58 |
| 7 | Mistral-7B-v0.1 | 83.08 |
| 8 | vicuna-33B-v1.3 | 81.59 |
| 9 | LLaMA-65B | 80.1 |
| 10 | LLaMA-30B | 79.1 |
| 11 | internlm-20B Chat | 78.11 |
| 12 | trurl-2-13B-academic | 78.11 |
| 13 | vicuna-13B-v1.5 | 77.61 |
| 14 | falcon-40B Instruct | 76.62 |
| 15 | Llama 2 13B Chat Base | 75.12 |
Interactive version: theaggregate.ai/benchmark?slug=mmlu-by-task-sociology · How the rankings work · Data refreshed daily, snapshot 2026-07-22.