RACE Attention is a strictly linear-time attention mechanism that approximates softmax attention outputs using Gaussian projections and soft LSH to enable training on contexts up to 12 million tokens.
Proof.(1)By definition,X (ℓ)= ˆS(ℓ)−E[ˆS(ℓ)], henceE[X (ℓ)] =E[ ˆS(ℓ)]−E[ˆS(ℓ)] = 0
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
RACE Attention: A Strictly Linear-Time Attention Layer for Training on Outrageously Large Contexts
RACE Attention is a strictly linear-time attention mechanism that approximates softmax attention outputs using Gaussian projections and soft LSH to enable training on contexts up to 12 million tokens.