Physical Diagram Generation on PhyDrawGen Mechanics (Evaluation Set)
65.3Preference RatePhyDrawGen
Evaluation Results
| Method | Links | |
|---|---|---|
| PhyDrawGenEvaluation Protocol=Human-judge forced-choice comparison2026.05 | 65.3 | |
| Gemini 3 ProEvaluation Protocol=Human-judge forced-choice comparison2026.05 | 18.8 | |
| GPT-5Evaluation Protocol=Human-judge forced-choice comparison2026.05 | 11.2 | |
| Gemini 2.5 FlashEvaluation Protocol=Human-judge forced-choice comparison2026.05 | 4.7 |