Model techniques map
Techniquesmodel architecturetoken mixersparse attention

specific method · filed under model architecture

Gated DeepSeek Sparse Attention

A gated variant of DeepSeek Sparse Attention; the evidence also mentions its use with IndexCache.

sources
3
model
1
lab adopt it
1
strongest
core

How sources treat it

One count per evidence span, weakest treatment to strongest.

core 3

Documented in

Further reading

Picked by hand, not extracted: where to read more, not evidence for anything on this page.

Evidence

3 spans quoted from the sources, strongest treatment first.

the attention module employs Gated DeepSeek Sparse Attention (Gated DSA)

coremodel architecturein Hy4-previewTencent Hunyuan

the attention module employs Gated DeepSeek Sparse Attention (Gated DSA) with IndexCache for cross-layer sparse index reuse

coremodel architecturein Hy4-previewTencent Hunyuan

the attention module employs Gated DeepSeek Sparse Attention (Gated DSA) with IndexCache

coremodel architecturein Hy4-previewTencent Hunyuan

Filed alongside

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