Physical Commonsense Reasoning on PIQA (ACC)
81.3ACCQwen2.5-VL
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-VLModel Category=General VLMs2026.07 | 81.3 | |
| ELSA3DModel Category=Unified 3D Models2026.07 | 80.6 | |
| LLaMA3.2-VisionModel Category=General VLMs2026.07 | 80.1 | |
| LLaMA-MeshModel Category=Mesh LLM2026.07 | 79.8 | |
| CoRe3DModel Category=Unified 3D Models2026.07 | 79.4 | |
| ShapeLLM-OmniModel Category=Unified 3D Models2026.07 | 78.9 | |
| ADAMWBackbone=Llama-500M, Evaluation Protocol=Zero-shot, Communication Strategy=DENSE all-reduce2026.07 | 67.3 | |
| ADAMWSynchronization=DENSE all-reduce, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 67.3 | |
| SCAPEBackbone=Llama-500M, Evaluation Protocol=Zero-shot, Sparsity (d)=0.12026.07 | 66.76 | |
| SCAPESparsity (d)=0.1, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 66.76 | |
| SCAPEBackbone=Llama-500M, Evaluation Protocol=Zero-shot, Sparsity (d)=0.012026.07 | 66.32 | |
| SCAPESparsity (d)=0.01, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 66.32 | |
| ADAMSBackbone=Llama-500M, Evaluation Protocol=Zero-shot, Communication Strategy=DENSE all-reduce2026.07 | 65.94 | |
| ADAMSSynchronization=DENSE all-reduce, Backbone=Llama-500M, Evaluation Protocol=Zero-shot2026.07 | 65.94 |