Empirical Characterization of Logging Smells in Machine Learning Code.

Patrick Loic Foalem, Leuson Da Silva, Foutse Khomh, Ettore Merlo, Heng Li{patrick-loic.foalem, leuson-mario-pedro.da-silva, foutse.khomh, ettore.merlo, heng.li}@polymtl.ca · 2026

Read Empirical Characterization of Logging Smells in Machine Learning Code. by Patrick Loic Foalem, Leuson Da Silva, Foutse Khomh, Ettore Merlo, Heng Li{patrick-loic.foalem, leuson-mario-pedro.da-silva, foutse.khomh, ettore.merlo, heng.li}@polymtl.ca (2026) 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-2601-05540.txt.

Description

Patrick Loic Foalem and 4 co-authors lay out, in a 2026 study plan, an empirical investigation of logging smells in machine learning code. Logging records what a system did, which supports monitoring, debugging, and auditing. In machine learning systems it also underpins reproducibility and traceability, yet recurring bad logging habits have never been catalogued systematically.

The planned method mines open-source machine learning repositories on GitHub to catalogue such smells. A survey of practicing engineers will then rate how relevant, severe, and frequent each smell feels. The authors note their findings may not generalize to closed industrial projects. Even so, the study aims to give the first evidence-based picture of logging practice in machine learning development.

Catalog record

CREATORPatrick Loic Foalem, Leuson Da Silva, Foutse Khomh, Ettore Merlo, Heng Li{patrick-loic.foalem, leuson-mario-pedro.da-silva, foutse.khomh, ettore.merlo, heng.li}@polymtl.ca
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE37 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-05540

Browse the full library · About the Tower of Babel