Model techniques map
Techniquesmodel architecturetoken mixersparse attention

implementation detail · filed under model architecture

Token-wise compression

A compression component mentioned alongside DeepSeek Sparse Attention, with no further mechanism specified.

source
1
models
2
lab adopt it
1
strongest
core

How sources treat it

One count per evidence span, weakest treatment to strongest.

core 1

Documented in

Evidence

1 span quoted from the sources, strongest treatment first.

Novel Attention: Token-wise compression + DSA (DeepSeek Sparse Attention)

coremodel architecturein DeepSeek-V4DeepSeek

Filed alongside

Other methods under model architecture :: token mixer :: sparse attention.