Style Alignment on Jane Austen (test)
0.984Mean Style ScoreFT-Model
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FT-ModelModel=GRPO fine-tuned 8B story generator2025.12 | 0.984 | 0.03 | |
| FT-AgenticEvaluation Protocol=Few-shot calibrated, Number of Parameters=8B2025.12 | 0.971 | — | |
| GPT-oss-120BEvaluation Protocol=Few-shot calibrated, Number of Parameters=120B2025.12 | 0.548 | — | |
| GPT-oss-120BModel=Reference model, Evaluation protocol=Zero-shot2025.12 | 0.521 | 0.11 | |
| Qwen2.5-32BModel=Open-weight baseline, Evaluation protocol=Zero-shot2025.12 | 0.422 | 0.11 | |
| Qwen2.5-32BEvaluation Protocol=Few-shot calibrated, Number of Parameters=32B2025.12 | 0.37 | — | |
| Llama3-70BModel=Open-weight baseline, Evaluation protocol=Zero-shot2025.12 | 0.356 | 0.083 | |
| Llama3.3-70BEvaluation Protocol=Few-shot calibrated, Number of Parameters=70B2025.12 | 0.356 | — | |
| Gemma3-27BModel=Open-weight baseline, Evaluation protocol=Zero-shot2025.12 | 0.283 | 0.086 | |
| Gemma3-27BEvaluation Protocol=Few-shot calibrated, Number of Parameters=27B2025.12 | 0.245 | — |