Abstract
The implementation of an embedded encoding of molecular representations reduces memory and disk usage while maintaining performance in deep learning models.
Funding Information
  • Google (GCP19980904)
  • Engineering and Physical Sciences Research Council (2602452, EP/T028513/1)
  • Royal Society of Chemistry (E21-7051491439, E21-8254227705)

This publication has 40 references indexed in Scilit: