@kalomaze we encode knowledge in language
language has some intrinsic mathematical structure
we learn these deep representations at scale
should suffice explaining this flavour of "AI"
For the past years my research focus was on unifying models and training paradigms across modalities. Today I'm excited that we're releasing our latest model aligned with this theme:
Gemma 4 12B, a dense encoder-free model which processes raw text, image, and audio inputs!
1/