Hybrid Mamba–Transformer Decoder for Error-Correcting Codes

Shy-el Cohen, Yoni Choukroun, Eliya Nachmani · 2025

Read Hybrid Mamba–Transformer Decoder for Error-Correcting Codes by Shy-el Cohen, Yoni Choukroun, Eliya Nachmani (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-17834.txt.

Description

Shy-el Cohen, Yoni Choukroun, and Eliya Nachmani build a hybrid decoder for error correcting codes that blends two neural designs, in a paper from 2025. One design reads sequences with great speed, while the other holds the full context of a code block.

Each layer of the fast design wears a mask that lets it heed only the code traits that matter at that depth. A staged loss checks the net at mid points too, which builds strong traits early. Across a range of linear codes, the hybrid beat both plain transformer decoders and plain fast design models by a clear margin.

Catalog record

CREATORShy-el Cohen, Yoni Choukroun, Eliya Nachmani
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-17834

Browse the full library · About the Tower of Babel