specific method · filed under model architecture
Causal Encoder-Decoder (CED) architecture
An architecture with a causal encoder followed by a decoder; the evidence also describes projecting decoder global key-value states from the final encoder hidden states.
Also called Causal Encoder-Decoder architecture, Causal Encoder-Decoder (CED).
- sources
- 4
- model
- 1
- lab adopt it
- 1
- strongest
- core
How sources treat it
One count per evidence span, weakest treatment to strongest.
Documented in
Further reading
Picked by hand, not extracted: where to read more, not evidence for anything on this page.
- Introducing DeepSeek-V4.1-Flash explainer www.deepseek.comintroduces the Causal Encoder-Decoder (CED) architecture
Evidence
7 spans quoted from the sources, strongest treatment first.
New Causal Encoder–Decoder architecture: just 8B active parameters for input, 16B for output.
We adopt a Causal Encoder-Decoder (CED) architecture, in which decoder global KV is projected from the final encoder hidden states.
The Causal Encoder–Decoder (CED) architecture and Compressed Sparse Attention 2 (CSA2) address complementary costs of long-context inference.
Its Causal Encoder-Decoder (CED) architecture enables the model to activate only 8B parameters per token during prefill, compared with 16B during decode, improving cost efficiency for input-heavy agentic workloads.
DeepSeek-V4.1-Flash adopts a Causal Encoder-Decoder (CED) architecture: a 40-layer Transformer organized as a 20-layer causal encoder followed by a 20-layer decoder.
the first built on the company's Causal Encoder-Decoder (CED) architecture
We adopt a Causal Encoder-Decoder architecture, with 20 layers in the encoder and 20 layers in the decoder.
Filed alongside
Other methods under model architecture.