Image-to-Text Retrieval on PhenoBench
63.3Recall@10PhenoLIP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| PhenoLIPEncoder (Vision)=ViT-B, Encoder (Text)=PubmedBert2026.02 | 63.3 | 81.92 | |
| BIOMEDICAEncoder (Vision)=ViT-L, Encoder (Text)=GPT22026.02 | 40.51 | 66.28 | |
| PMC-CLIPEncoder (Vision)=ViT-L, Encoder (Text)=PubmedBert2026.02 | 40 | 64.82 | |
| BiomedCLIPEncoder (Vision)=ViT-B, Encoder (Text)=PubmedBert2026.02 | 32.91 | 56.63 | |
| SigLIP2Encoder (Vision)=So400m, Encoder (Text)=SigLIP642026.02 | 14.02 | 28.16 | |
| PhenoLIPEncoder (Vision)=ViT-B, Encoder (Text)=PubmedBert2026.02 | 13.84 | 36.88 | |
| OpenCLIPEncoder (Vision)=ViT-B, Encoder (Text)=GPT22026.02 | 10.72 | 24.25 | |
| CoCaEncoder (Vision)=ViT-B, Encoder (Text)=GPT22026.02 | 9.27 | 22.16 | |
| BIOMEDICAEncoder (Vision)=ViT-L, Encoder (Text)=GPT22026.02 | 8.12 | 25.27 | |
| PMC-CLIPEncoder (Vision)=ViT-L, Encoder (Text)=PubmedBert2026.02 | 7.22 | 23.44 | |
| BiomedCLIPEncoder (Vision)=ViT-B, Encoder (Text)=PubmedBert2026.02 | 3.71 | 13.38 | |
| OpenCLIPEncoder (Vision)=ViT-B, Encoder (Text)=GPT22026.02 | 2.88 | 8.34 | |
| CoCaEncoder (Vision)=ViT-B, Encoder (Text)=GPT22026.02 | 2.02 | 6.42 | |
| SigLIP2Encoder (Vision)=So400m, Encoder (Text)=SigLIP642026.02 | 0.31 | 0.76 |