Controllable Narrative Rendering on ROCStories (test)
2.83RPB ScoreLoom
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| LoomModel Paradigm=Thinking Model2026.05 | 2.83 | 4.83 | 3.93 | 3.86 | |
| Qwen2.5-7BModel Paradigm=Language Model2026.05 | 1.04 | 2.45 | 3.39 | 2.29 | |
| o4-miniModel Paradigm=Thinking Model2026.05 | 1.04 | 3.96 | 4.31 | 3.1 | |
| Qwen2.5-32BModel Paradigm=Language Model2026.05 | 1.02 | 3.48 | 3.84 | 2.78 | |
| Qwen3-235BModel Paradigm=Thinking Model2026.05 | 1.02 | 3.78 | 3.96 | 2.92 | |
| GPT-5-thinkingModel Paradigm=Thinking Model2026.05 | 1.02 | 4.29 | 4.33 | 3.21 | |
| Qwen2.5-14BModel Paradigm=Language Model2026.05 | 1.01 | 2.79 | 3.55 | 2.45 |