Open Models Almanac
Methodsmodel architecturemultimodal architecture

implementation detail · filed under model architecture

Raw audio projection into LLM embedding space

Projects raw audio chunks or signals directly into the language-model or text-token embedding space without a separate audio encoder.

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 raw audio chunk projection.

sources
2
model
1
lab adopt it
1
strongest
used

Documented in

Google DeepMind

Evidence

2 quotes from 2 sources, one card per model, each opening with its strongest.

How sources treat it
used 2

Gemma 4 31B

Google DeepMindadopted2 sources

Raw audio is segmented into 40ms chunks at 16kHz, resulting in 640-dimensional vectors per chunk. These are projected directly into the LLM embedding space.

usedmodel architecturein Gemma 4 12BGoogle DeepMind
1 more from this model

We removed the audio encoder entirely and projected the raw audio signal into the same dimensional space as text tokens.

usedmodel architecturein Gemma 4 12BGoogle
↑