Binary comparison for commonsense plausibility on ViComTe Color 1.0 (test)
Evaluation results
| Method | Links | |
|---|
| GPT-42025.02 | | 93.29 |
|---|
| EVA-CLIPFramework=ComPaSS2025.02 | | 93.29 |
|---|
| GPT-3.52025.02 | | 92.25 |
|---|
| Qwen2Framework=ComPaSS, Training=Contrastive Learning (CL)2025.02 | | 86.79 |
|---|
| MistralFramework=ComPaSS, Training=Contrastive Learning (CL)2025.02 | | 86.06 |
|---|