Scene Synthesis on Large-Cloud Computing Scale (test)
1.145PUEPhythesis (Full)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Phythesis (Full)Type=Iterative, Backbone=R-LM2025.12 | 1.145 | 0.97 | |
| ABL (w/o phy)Type=Iterative, Backbone=R-LM2025.12 | 1.151 | 0.893 | |
| ABL (w/o reflect, phy)Type=Iterative, Backbone=R-LM2025.12 | 1.152 | 0.697 | |
| Vanilla LLMType=Direct, Backbone=R-LM2025.12 | 1.162 | 0.813 | |
| ABL (w/o design, phy)Type=Iterative, Backbone=R-LM2025.12 | 1.164 | 0.677 | |
| Phythesis (Full)Type=Iterative, Backbone=I-LM2025.12 | 1.175 | 0.931 | |
| ABL (w/o phy)Type=Iterative, Backbone=I-LM2025.12 | 1.193 | 0.729 | |
| ABL (w/o llm)Type=Iterative, Backbone=I-LM2025.12 | 1.206 | 0.557 | |
| ABL (w/o design, phy)Type=Iterative, Backbone=I-LM2025.12 | 1.23 | 0.422 | |
| EAType=Iterative, Backbone=I-LM2025.12 | 1.246 | 0.63 | |
| ABL (w/o reflect, phy)Type=Iterative, Backbone=I-LM2025.12 | 1.274 | 0.679 | |
| RandomType=Direct, Backbone=I-LM2025.12 | 1.302 | 0.429 | |
| Vanilla LLMType=Direct, Backbone=I-LM2025.12 | 1.366 | 0.36 |