Cognition and Reasoning on OKVQA
61.92ScoreTask Arithmetic
Evaluation results
| Method | Links | |
|---|---|---|
| Task ArithmeticMerging Protocol=Task Arithmetic2025.03 | 61.92 | |
| AdaMMSMerging Protocol=AdaMMS2025.03 | 61.11 | |
| DARE-LinearMerging Protocol=DARE-Linear2025.03 | 60.96 | |
| MetaGPTMerging Protocol=MetaGPT2025.03 | 60.84 | |
| CogVLM-chat-7BStatus=Base Model2025.03 | 60.82 | |
| AdaMMSmerging_target=Qwen2-VL-7B, merging_source=LLaVA-OneVision-7B2025.03 | 53.56 | |
| LLaVA-v1.5-7BStatus=Source Model2025.03 | 53.42 | |
| DARE-TiesMerging Protocol=DARE-Ties2025.03 | 53.15 | |
| Ties-Mergingmerging_target=Qwen2-VL-7B, merging_source=LLaVA-OneVision-7B2025.03 | 52.56 | |
| MetaGPTmerging_target=Qwen2-VL-7B, merging_source=LLaVA-OneVision-7B2025.03 | 52.24 | |
| Qwen2-VL (base)model_status=Original Model2025.03 | 51.43 | |
| Task Arithmeticmerging_target=Qwen2-VL-7B, merging_source=LLaVA-OneVision-7B2025.03 | 50.6 | |
| DARE-Tiesmerging_target=Qwen2-VL-7B, merging_source=LLaVA-OneVision-7B2025.03 | 49.61 | |
| LLaVA-OneVisionmodel_status=Original Model2025.03 | 49.57 | |
| DARE-Linearmerging_target=Qwen2-VL-7B, merging_source=LLaVA-OneVision-7B2025.03 | 43.41 | |
| Ties-MergingMerging Protocol=Ties-Merging2025.03 | 40.31 |