RAIDIUM

Multimodality means the model can handle several types of inputs (text, image, sound etc). For example, the latest version of ChatGPT can support both images and texts... They are trained on very large general datasets, with a method called Self-Supervised Learning (SSL). Contrary to classic supervised models, they are not limited to one task, and can perform a bunch of downstream medical analysis (imaging biomarkers)... Radiology augmented GPT to tackle the complexity of precision medicine

VAT numbers: FR66909501991

Primary location: Aubervilliers France