Multidomain Expert Ensembles for Robust AI-Generated Text Detection With Backdoor-Aware Continual Learning
Computer · 2026-07 · 59 (7) · 22-33
DOI: 10.1109/mc.2026.3681589 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computer, 2026-07. Verfasst von Ian Miller und 1 weiteren Autor:innen. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
We present a system to detect artificial intelligence-generated text. Our approach combines multiple expert detectors with a domain classifier that routes incoming text to the most appropriate expert. Results show that our ensemble outperforms individual detectors, both on average and within each domain category.
Zitieren
APA
Miller, I., & Lin, D. (2026). Multidomain Expert Ensembles for Robust AI-Generated Text Detection With Backdoor-Aware Continual Learning. Computer, 59(7), 22-33. https://doi.org/10.1109/mc.2026.3681589
Vancouver
Miller I, Lin D. Multidomain Expert Ensembles for Robust AI-Generated Text Detection With Backdoor-Aware Continual Learning. Computer. 2026;59(7):22-33. doi:10.1109/mc.2026.3681589
BibTeX
@article{Miller2026,
author = {Miller, Ian and Lin, Dan},
title = {Multidomain Expert Ensembles for Robust AI-Generated Text Detection With Backdoor-Aware Continual Learning},
journal = {Computer},
year = {2026},
volume = {59},
number = {7},
pages = {22--33},
doi = {10.1109/mc.2026.3681589}
}
EUVIMED