The vision encoder and connector turn visual inputs into representations processed by the language backbone.
1 more from this model
A C-RADIO vision encoder and connector extend the language backbone to visual inputs
specific method · filed under model architecture
Uses a C-RADIO vision encoder to provide visual representations for the language backbone.
The sources on this page name it without explaining how it works. The line above is the extraction's one-line gloss. See multimodal architecture for the methods filed beside it.
2 quotes from 2 sources, one card per model, each opening with its strongest.
The vision encoder and connector turn visual inputs into representations processed by the language backbone.
A C-RADIO vision encoder and connector extend the language backbone to visual inputs