Softmax attention has algebraic invariants including zero-sum rows and head-dimension rank limits, plus consistent variance spread in language models attributed to key incoherence.
Academic Press, 3rd edition, 2009
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
On the Invariants of Softmax Attention
Softmax attention has algebraic invariants including zero-sum rows and head-dimension rank limits, plus consistent variance spread in language models attributed to key incoherence.