HELM MMLU - Formal Logic: leaderboard
The "Formal Logic" subtask of HELM MMLU. HELM Classic MMLU group leaderboard from the public HELM release artifacts.
Metric: EM. Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 79 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet (20241022) | 78.57 | #291 |
| 2 | DeepSeek V3 | 76.98 | #323 |
| 3 | Gemini 1.5 Pro (002) | 75.4 | #350 |
| 4 | Qwen 2.5 72B Instruct | 73.02 | #458 |
| 5 | Nova Pro | 71.43 | #573 |
| 6 | GPT-4 Turbo | 70.63 | #406 |
| 7 | Gemini 1.5 Pro (001) | 70.63 | #446 |
| 8 | Claude 3.5 Sonnet (20240620) | 69.84 | #298 |
| 9 | Llama 3.1 405B Instruct | 69.84 | #461 |
| 10 | Claude 3 Opus (20240229) | 69.84 | #413 |
| 11 | GPT-4o (2024-05-13) | 68.25 | #326 |
| 12 | Llama 3.2 90B Vision Instruct | 68.25 | #626 |
| 13 | Llama 3.3 70B Instruct | 67.46 | #536 |
| 14 | Llama 3.1 70B Instruct | 67.46 | #566 |
| 15 | GPT-4o (2024-08-06) | 67.46 | #335 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-mmlu-formal-logic · How It Works · Data refreshed daily, snapshot 2026-10-23.