Model techniques map
Techniquesmodel architecturechannel mixerdense feed-forward network

implementation detail · filed under model architecture

Dense feed-forward network

A dense FFN used in the MTP block instead of a mixture-of-experts block.

Also called dense Feed-Forward Network (FFN).

source
1
model
1
lab adopt it
1
strongest
used

How sources treat it

One count per evidence span, weakest treatment to strongest.

used 1

Documented in

Evidence

1 span quoted from the sources, strongest treatment first.

MTP block uses dense FFN instead of MoE

usedmodel architecturein MiMo-V2-Flash (MTP block)Xiaomi

Filed alongside

Other methods under model architecture :: channel mixer :: dense feed-forward network.