18 citations · 18 across the 1 of their papers we have counts for
1 paper
Chameleon Team
We present Chameleon, a family of early-fusion token-based mixed-modal models capable of understanding and generating images and text in any arbitrary sequence. We outline a stable…