STEMVerse - Physics - Electrodynamics: leaderboard
Metric: Few-shot accuracy (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen3-14B-Base | 36.61 |
| 2 | Qwen3-8B-Base | 28.87 |
| 3 | Llama 3.1 8B Instruct | 22.95 |
| 4 | Qwen 2.5 14B Instruct | 20.95 |
| 5 | Qwen 2.5 14B | 20.31 |
| 6 | Llama 3.1 8B | 19.95 |
| 7 | Qwen 2.5 7B | 18.85 |
| 8 | Qwen 2.5 3B | 18.4 |
| 9 | Llama 3.2 3B Instruct | 16.48 |
| 10 | Qwen 2.5 7B Instruct | 16.03 |
| 11 | Qwen 2.5 3B Instruct | 15.27 |
| 12 | Llama 3.2 3B | 15.21 |
| 13 | Qwen3-4B-Base | 13.21 |
Interactive version: theaggregate.ai/benchmark?slug=stemverse-physics-electrodynamics · How It Works · Data refreshed daily, snapshot 2026-09-25.