implementation detail · filed under model architecture
Fixed Sliding Window Attention
A sliding-window attention baseline described as FLOP-matched; its window size is not specified.
Also called sliding-window baseline.
- source
- 1
- model
- 1
- labs adopt it
- 0
- strongest
- evaluated
How sources treat it
One count per evidence span, weakest treatment to strongest.
evaluated 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
a FLOP-matched sliding-window baseline
evaluatedmodel architecturein MiniMax Sparse AttentionMiniMax
Filed alongside
Other methods under model architecture :: token mixer :: softmax attention :: sliding window attention.
Sliding Window AttentionHybrid Sliding Window Attention512-token Sliding Window AttentionDecoder SWA Bounded ReplayAlternating Row-Major and Column-Major Token SerializationDynamic Attention Window Size TrainingPer-Head GatingPure Sliding Window AttentionSink-Augmented Sliding Window AttentionSWA-1024SWA-128