Multi-modal Representation Learning on MMEB Overall 1.0
61.6Classification P@1SDE
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SDEEvaluation Protocol=Fine-tuning, Model Backbone=LLaVA-1.6, Resolution=1344x1344 (HR)2026.02 | 61.6 | 54.7 | 69 | 92.5 | 65.6 | |
| VLM2VecEvaluation Protocol=Fine-tuning, Model Backbone=LLaVA-1.6, Resolution=1344x1344 (HR)2026.02 | 61.2 | 49.9 | 67.4 | 86.1 | 62.9 | |
| UniMEEvaluation Protocol=Fine-tuning, Model Backbone=LLaVA-1.62026.02 | 60.6 | 52.9 | 67.9 | 85.1 | 63.6 | |
| SDEEvaluation Protocol=Fine-tuning, Model Backbone=Qwen2-VL-2B2026.02 | 60 | 51 | 66.4 | 72.5 | 61.1 | |
| VLM2VecEvaluation Protocol=Fine-tuning, Model Backbone=Qwen2-VL-2B2026.02 | 59 | 49.4 | 65.4 | 73.4 | 60.1 | |
| SDEEvaluation Protocol=Fine-tuning, Model Backbone=LLaVA-1.6, Resolution=336x336 (LR)2026.02 | 57.2 | 43.6 | 63.6 | 87.7 | 59 | |
| OpenCLIP-FFTEvaluation Protocol=Fine-tuning2026.02 | 56 | 21.9 | 55.4 | 64.1 | 47.2 | |
| MegaPairsEvaluation Protocol=Fine-tuning, Model Backbone=LLaVA-1.62026.02 | 56 | 57.4 | 69.9 | 83.6 | 64.1 | |
| CLIP-FFTEvaluation Protocol=Fine-tuning2026.02 | 55.2 | 19.7 | 53.2 | 62.2 | 45.4 | |
| VLM2VecEvaluation Protocol=Fine-tuning, Model Backbone=Phi3.5-V2026.02 | 54.8 | 54.9 | 62.3 | 79.5 | 60.1 | |
| VLM2VecEvaluation Protocol=Fine-tuning, Model Backbone=LLaVA-1.6, Resolution=336x336 (LR)2026.02 | 54.7 | 50.3 | 56.2 | 64 | 55 | |
| UniMEEvaluation Protocol=Fine-tuning, Model Backbone=Phi3.5-V2026.02 | 54.6 | 55.9 | 64.5 | 81.8 | 61.3 | |
| SDEEvaluation Protocol=Fine-tuning, Model Backbone=Phi3.5-V2026.02 | 54.6 | 60.3 | 63.4 | 78.4 | 61.7 | |
| UniIREvaluation Protocol=Zero-shot, Fusion Strategy=score-level, Backbone=CLIP2026.02 | 44.3 | 16.2 | 61.8 | 65.3 | 44.7 | |
| CLIPEvaluation Protocol=Zero-shot2026.02 | 42.8 | 9.1 | 53 | 51.8 | 37.8 | |
| UniIREvaluation Protocol=Zero-shot, Fusion Strategy=feature-level, Backbone=BLIP2026.02 | 42.1 | 15 | 60.1 | 62.2 | 42.8 | |
| MagiclensEvaluation Protocol=Zero-shot2026.02 | 38.8 | 8.3 | 35.4 | 26 | 27.8 | |
| BLIP2Evaluation Protocol=Zero-shot2026.02 | 27 | 4.2 | 33.9 | 47 | 25.2 |