2 papers
cs.LG2026
MXNorm: Reusing MXFP block scales for efficient tensor normalisation
Callum McLean, Luke Y. Prince, Alexandre Payot +2
Matrix multiplication performance has long been the major bottleneck to scaling deep learning workloads, which has stimulated the design of new accelerators that use increasingly l…
cs.LG2024
: A Parameter-Efficient Foundation Model for Molecular Learning
Kerstin Kläser, BÅażej Banaszewski, Samuel Maddrell-Mander +5
In biological tasks, data is rarely plentiful as it is generated from hard-to-gather measurements. Therefore, pre-training foundation models on large quantities of available data a…