Text Generation on WritingPrompts
22.11F1 ScorePrompt-R1
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Prompt-R1Backbone=GPT-4o-mini2025.11 | 22.11 | 38.54 | |
| TextGradOptimization Framework=APO, Backbone=GPT-4o-mini2025.11 | 22.07 | 20.28 | |
| BaselineBackbone=GPT-4o-mini2025.11 | 19.48 | 35.31 | |
| SFTBackbone=Qwen3-4B2025.11 | 15.12 | 30.86 | |
| BaselineBackbone=Qwen3-4B2025.11 | 13.06 | 20.59 | |
| CoT ReasoningBackbone=Qwen3-4B2025.11 | 10.2 | 12.15 | |
| CoT ReasoningBackbone=GPT-4o-mini2025.11 | 9.93 | 30.86 | |
| GRPOBackbone=Qwen3-4B2025.11 | 8.4 | 10.41 | |
| OPROOptimization Framework=APO, Backbone=GPT-4o-mini2025.11 | 7.81 | 13.73 | |
| GEPAOptimization Framework=APO, Backbone=GPT-4o-mini2025.11 | 0.21 | 0.34 |