Explainable Disentangled Representation Learning for Generalizable Authorship Attribution in the Era of Generative AI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Man, Hieu, Pham, Van-Cuong, Ngo, Nghia Trung, Dernoncourt, Franck, Nguyen, Thien Huu
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911617104478208
author Man, Hieu
Pham, Van-Cuong
Ngo, Nghia Trung
Dernoncourt, Franck
Nguyen, Thien Huu
author_facet Man, Hieu
Pham, Van-Cuong
Ngo, Nghia Trung
Dernoncourt, Franck
Nguyen, Thien Huu
contents Learning robust representations of authorial style is crucial for authorship attribution and AI-generated text detection. However, existing methods often struggle with content-style entanglement, where models learn spurious correlations between authors' writing styles and topics, leading to poor generalization across domains. To address this challenge, we propose Explainable Authorship Variational Autoencoder (EAVAE), a novel framework that explicitly disentangles style from content through architectural separation-by-design. EAVAE first pretrains style encoders using supervised contrastive learning on diverse authorship data, then finetunes with a Variational Autoencoder (VEA) architecture using separate encoders for style and content representations. Disentanglement is enforced through a novel discriminator that not only distinguishes whether pairs of style/content representations belong to the same or different authors/content sources, but also generates natural language explanation for their decision, simultaneously mitigating confounding information and enhancing interpretability. Extensive experiments demonstrate the effectiveness of EAVAE. On authorship attribution, we achieve state-of-the-art performance on various datasets, including Amazon Reviews, PAN21, and HRS. For AI-generated text detection, EAVAE excels in few-shot learning over the M4 dataset. Code and data repositories are available online\footnote{https://github.com/hieum98/avae} \footnote{https://huggingface.co/collections/Hieuman/document-level-authorship-datasets}.
format Preprint
id arxiv_https___arxiv_org_abs_2604_21300
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Explainable Disentangled Representation Learning for Generalizable Authorship Attribution in the Era of Generative AI
Man, Hieu
Pham, Van-Cuong
Ngo, Nghia Trung
Dernoncourt, Franck
Nguyen, Thien Huu
Computation and Language
Information Retrieval
Machine Learning
Learning robust representations of authorial style is crucial for authorship attribution and AI-generated text detection. However, existing methods often struggle with content-style entanglement, where models learn spurious correlations between authors' writing styles and topics, leading to poor generalization across domains. To address this challenge, we propose Explainable Authorship Variational Autoencoder (EAVAE), a novel framework that explicitly disentangles style from content through architectural separation-by-design. EAVAE first pretrains style encoders using supervised contrastive learning on diverse authorship data, then finetunes with a Variational Autoencoder (VEA) architecture using separate encoders for style and content representations. Disentanglement is enforced through a novel discriminator that not only distinguishes whether pairs of style/content representations belong to the same or different authors/content sources, but also generates natural language explanation for their decision, simultaneously mitigating confounding information and enhancing interpretability. Extensive experiments demonstrate the effectiveness of EAVAE. On authorship attribution, we achieve state-of-the-art performance on various datasets, including Amazon Reviews, PAN21, and HRS. For AI-generated text detection, EAVAE excels in few-shot learning over the M4 dataset. Code and data repositories are available online\footnote{https://github.com/hieum98/avae} \footnote{https://huggingface.co/collections/Hieuman/document-level-authorship-datasets}.
title Explainable Disentangled Representation Learning for Generalizable Authorship Attribution in the Era of Generative AI
topic Computation and Language
Information Retrieval
Machine Learning
url https://arxiv.org/abs/2604.21300