We propose a new approach for the authorship attribution task that leverages the various linguistic representations learned at different layers of pre-trained transformer-based models. We evaluate our approach on three datasets, comparing it to a state-of-the-art baseline in in-domain and out-of-domain scenarios. We found that utilizing various transformer layers improves the robustness of authorship attribution models when tested on out-of-domain data, resulting in new state-of-the-art results. Our analysis gives further insights into how our model's different layers get specialized in representing certain stylistic features that benefit the model when tested out of the domain.
@article{arxiv.2503.00958,
title = {Layered Insights: Generalizable Analysis of Authorial Style by Leveraging All Transformer Layers},
author = {Milad Alshomary and Nikhil Reddy Varimalla and Vishal Anand and Smaranda Muresan and Kathleen McKeown},
journal= {arXiv preprint arXiv:2503.00958},
year = {2025}
}