Explore Log in Sign up

Explore

@jmcastagnetto.bsky.social Oct 7, 2026, 9:28 PM

From #Nature: "An open vision-language model for diverse #medical applications" www.nature.com/articles/s41...

#AI #Medgemma #LLM #ComputerVision

The MedGemma model collection features the MedSigLIP image encoder (400 million parameters), MedGemma 4B (4 billion parameters, multimodal), MedGemma 27B (27 billion parameters, multimodal) and MedGemma 27B Text (27 billion parameters, text only). Modalities used for domain-specific training of the respective models are indicated by the lines connecting data types on the left to models on the right. EHR, electronic health record.

Terms of Use Privacy Policy Community Standards Help Get the app

Grilled Cheese is a product of Village Compute

Version devBuilt at: 2026-10-11 19:07:23 EDT