Model techniques map
Techniquesmodel architecturecontext capacity

specific method · filed under model architecture

Engram

A sparsely accessed conditional memory module that separates memorization from computation.

Also called Engram conditional memory, Engram conditional memory module, sparsely accessed conditional memory.

sources
3
models
2
lab adopt it
1
strongest
core

How sources treat it

One count per evidence span, weakest treatment to strongest.

used 2core 2not used 1

Documented in

Further reading

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

Evidence

5 spans quoted from the sources, strongest treatment first.

Engram (Cheng et al., 2026b) adds sparsely accessed conditional memory.

coremodel architecturein DeepSeek-V4.1-FlashDeepSeek

We augment DeepSeek-V4.1-Flash with Engram (Cheng et al., 2026c), the conditional memory module introduced in our previous work to decouple memorization from computation.

coreunclearin DeepSeek-V4.1-FlashDeepSeek

we integrate the Engram (Cheng et al., 2026b) conditional memory module to strengthen model capabilities.

usedmodel architecturein DeepSeek-V4.1-FlashDeepSeek

Engram conditional memory (196B parameters, sparsely accessed via token-based lookup)

usedmodel architecturein DeepSeek-V4.1-FlashDeepSeek

Engram conditional memory is absent from V4

not usedmodel architecturein DeepSeek-V4DeepSeek

Filed alongside

Other methods under model architecture :: context capacity.