Open Models Almanac
Methodsmodel architecturetoken mixersparse attentionsparse attention indexer

single mention · model architecture

Top-2,048 key-token selection per query

Selects up to 2,048 key tokens for each query using a lightweight indexer.

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

Filed under model architecture › token mixer › sparse attention › sparse attention indexer