Model techniques map
Techniquesmodel architecturenormalization & residual

specific method · filed under model architecture

RMSNorm

A normalization method based on the root mean square of activations; one cited use applies it to attention queries and compressed key-value entries.

Also called RMSNorm normalization, root mean square normalization (RMSNorm), root mean square normalization.

sources
3
models
4
labs adopt it
2
strongest
used

How sources treat it

One count per evidence span, weakest treatment to strongest.

used 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.

applying root mean square normalization [6] on the activations before each attention and MoE block

usedmodel architecturein gpt-oss-120b and gpt-oss-20bOpenAI

we perform an additional RMSNorm operation on each head of the queries and the only head of the compressed KV entries

usedmodel architecturein DeepSeek-V4DeepSeek

We also adopt RMSNorm (Zhang and Sennrich, 2019) for normalization

usedmodel architecturein DeepSeek-V4.1-FlashDeepSeek

Filed alongside

Other methods under model architecture :: normalization & residual.