Sign in

AtMan: Understanding Transformer Predictions Through Memory Efficient Attention Manipulation.

Mayukh DebBjörn DeiserothSamuel WeinbachPatrick SchramowskiKristian Kersting
Published in: CoRR (2023)
Keyphrases
  • memory efficient
  • external memory
  • iterative deepening
  • neural network
  • fuzzy logic
  • integral image
  • information retrieval
  • data structure
  • visual attention
  • deeper understanding
  • multiple sequence alignment
  • high voltage