Open Models Almanac
Methodsmodel architecturemultimodal architecture

single mention · model architecture

Patch-as-token vision-language architecture

Treats pretrained visual-transformer patch features as tokens and projects them into the language-model input space.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See multimodal architecture for the methods filed beside it.

Also called patch-as-token.

Filed under model architecture › multimodal architecture