AGF replaces softmax attention with a learned singular-value-domain graph filter that runs in O(n d^2), and reports moderate accuracy improvements on UEA and LRA benchmarks.
Transformers are rnns: Fast autoregressive transformers with linear attention
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Learning Advanced Self-Attention for Linear Transformers in the Singular Value Domain
AGF replaces softmax attention with a learned singular-value-domain graph filter that runs in O(n d^2), and reports moderate accuracy improvements on UEA and LRA benchmarks.