Quantizable Transformers: Removing Outliers by Helping Attention Heads Do Nothing.

Yelysei Bondarenko Markus Nagel Tijmen Blankevoort

Published in: NeurIPS (2023)

Keyphrases

outlier detection
data mining
mobile robot
missing data
visual attention
databases
neural network
information systems
bayesian networks
data streams
feature space
relational databases
robust regression