HELM Classic - Synthetic Reasoning Abstract — leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 69 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | code-davinci-002 | 53.96 |
| 2 | GPT-3.5 Turbo (0613) | 50.94 |
| 3 | text-davinci-003 | 50.25 |
| 4 | Llama 2 70B | 48.87 |
| 5 | text-davinci-002 | 48.78 |
| 6 | Mistral-7B-v0.1 | 45.11 |
| 7 | GPT-3.5 Turbo (0301) | 44.98 |
| 8 | LLaMA-65B | 43.82 |
| 9 | LLaMA-30B | 41.23 |
| 10 | mpt-30B | 33.53 |
| 11 | Llama 2 13B | 32.36 |
| 12 | falcon-40B | 23.75 |
| 13 | davinci | 23.58 |
| 14 | Llama 2 7B | 22.2 |
| 15 | LLaMA-13B | 21.42 |
Interactive version: theaggregate.ai/benchmark?slug=helm-classic-synthetic-reasoning-abstract · How the rankings work · Data refreshed daily, snapshot 2026-07-22.