Transition Prediction on WoW
41.32IoUOpus 4.6
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Opus 4.6Type=Frontier Models, Access to Business Rules (BR)=w/o BR2026.05 | 41.32 | 44.9 | |
| Sonnet 4.6Type=Frontier Models, Access to Business Rules (BR)=w/o BR2026.05 | 38.65 | 41.54 | |
| Gemini 3 ProType=Frontier Models, Access to Business Rules (BR)=w/o BR2026.05 | 35.59 | 39.06 | |
| Qwen-3.6-27B-LoRAType=Finetuned Models, Access to Business Rules (BR)=w/o BR2026.05 | 32.22 | 36.88 | |
| Gemma-4-31B-LoRAType=Finetuned Models, Access to Business Rules (BR)=w/o BR2026.05 | 31.73 | 36.07 | |
| Qwen-3.5-27B-LoRAType=Finetuned Models, Access to Business Rules (BR)=w/o BR2026.05 | 31.21 | 35.02 | |
| GPT-5Type=Frontier Models, Access to Business Rules (BR)=w/o BR2026.05 | 29.34 | 32.32 | |
| Qwen-3.6-27BType=Base Models, Access to Business Rules (BR)=w/o BR2026.05 | 23.17 | 27.02 | |
| Qwen-3.5-27BType=Base Models, Access to Business Rules (BR)=w/o BR2026.05 | 21.22 | 15.19 | |
| Gemma-4-31BType=Base Models, Access to Business Rules (BR)=w/o BR2026.05 | 21.11 | 23.55 |