Compressing Below the IID Shannon Bound

Zenodo (CERN European Organization for Nuclear Research) 2026 Open access

Read full text
Authors: Michael Ricky Neal
DOI: 10.5281/zenodo.22737445
Cited by: 2

Abstract

We describe a nine-to-eleven stage lossless compression pipeline for neural-network weighttensors that achieves output sizes below the classical (IID)

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
13 September 2026
Language
EN