Performers approximate full-rank softmax attention in Transformers via FAVOR+ random features for linear complexity, with theoretical guarantees of unbiased estimation and competitive results on pixel, text, and protein tasks.
Random Struct
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
ULPT optimizes prompts in ultra-low dimensions with frozen random up-projection to cut training parameters by 98% while matching vanilla prompt tuning performance on NLP tasks.
citing papers explorer
-
Rethinking Attention with Performers
Performers approximate full-rank softmax attention in Transformers via FAVOR+ random features for linear complexity, with theoretical guarantees of unbiased estimation and competitive results on pixel, text, and protein tasks.
-
Ultra-Low-Dimensional Prompt Tuning via Random Projection
ULPT optimizes prompts in ultra-low dimensions with frozen random up-projection to cut training parameters by 98% while matching vanilla prompt tuning performance on NLP tasks.