specific method · filed under model architecture
Two-stage introduction of sparse attention
A training strategy that introduces sparse attention through two stages; the evidence does not specify the stage operations.
Also called two-stage training method.
- source
- 1
- models
- 2
- 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.
the strategy of introducing sparse attention is the same as DeepSeek-V4-Flash, following a two-stage training method.
usedtraining objectivein DeepSeek-V4DeepSeek
Filed alongside
Other methods under model architecture :: token mixer :: sparse attention.
DeepSeek Sparse AttentionCompressed Sparse AttentionQwen Sparse AttentionSparse attentionHeavily Compressed AttentionGated DeepSeek Sparse AttentionCSA2 Full ModeKV-outer sparse attentionProgressive sequence-length extension for sparse attentionSparse-attention continued pre-training with joint model and indexer optimizationCross-layer KV and index reuse with statically assigned CSA2 modesFixed-budget sparse-attention selectionFrom-scratch sparse attention training without dense warmupJoint backbone and indexer training under sparse attentionNative Sparse AttentionNatively trained sparsityNoPE sparse multi-head latent attentionQSA micro-block compression at ratio 4Sequential block processingSparse retrieval over long contextsSparse softmax attentionToken-wise compressionTwo-stage sparse attention