Scene-Level Comparative Evaluation on Narrative Script Refinement (test)
90MeanDramaturge
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Dramaturge2025.10 | 90 | 0.49 | |
| Gemini-2.5-pro2025.10 | 76 | 0.94 | |
| Deepseek-r12025.10 | 63.96 | 0.94 | |
| GPT-4.12025.10 | 62.72 | 1.87 | |
| GPT-4.1-mini2025.10 | 61.7 | 0.97 | |
| Original Scripts2025.10 | 54.5 | 1.6 | |
| Qwen-max2025.10 | 52.36 | 0.78 | |
| GPT-4o2025.10 | 52.2 | 0.85 |