An Empirical Investigation of Pre-Trained Deep Learning Model Reuse in the Scientific Process

arXiv:2603.13584 · cs.SE, cs.AI · Submitted 2026-03-13 · Read on arXiv

cs.SE, cs.AI

Submitted: 2026-03-13

Updated: 2026-08-28

Comments: 24 pages (6 pages prose, 6 pages citations, 12 pages appendix), 9 figures, 2 tables, accepted as a short paper to the 22nd IEEE International Conference on eScience

Code: https://github.com/jgm/pandoc

License: http://creativecommons.org/licenses/by/4.0/

The gist: Deep learning has achieved recognition for its impact within natural sciences, yet the prohibitive financial and technical cost of training models from scratch inhibit adoption.

Terminology

Abstract

Deep learning has achieved recognition for its impact within natural sciences, yet the prohibitive financial and technical cost of training models from scratch inhibit adoption. Following software engineering community guidance, natural scientists are reusing pre-trained deep learning models (PTMs) to amortize these costs. While prior works recommend PTM reuse patterns, we present the first empirical study of PTM reuse patterns in the natural sciences, quantifying the utilization and impact of PTM reuse within the scientific process across 17,718 peer reviewed, open access papers. Our results show that "Biochemistry, Genetics and Molecular Biology" has outpaced other natural scientific fields in PTM reuse, "adaptation" reuse is the most prevalent PTM reuse pattern identified across all natural science fields, and the "testing" stage of the scientific process has been most impacted by PTM integration.

Sources

Related papers