@inproceedings{kulumba-etal-2026-authorship,
title = "Where Does Authorship Signal Emerge in Encoder-Based Language Models?",
author = "Kulumba, Francis and
Vimont, Guillaume and
Romary, Laurent and
Cafiero, Florian",
editor = "Arad, Dana and
Belinkov, Yonatan and
Chen, Hanjie and
Feldhus, Nils and
Mueller, Aaron and
Mohebbi, Hosein and
Sarti, Gabriele and
Tutek, Martin",
booktitle = "Proceedings of the 9th {B}lackbox{NLP} Workshop: Analyzing and Interpreting Neural Networks for {NLP}",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.blackboxnlp-1.47/",
pages = "818--834",
ISBN = "979-8-89176-476-7",
abstract = "Authorship attribution asks whether two texts were written by the same author. Models fine-tuned for this task with the same pretrained encoder, data, and loss can still differ four-fold in Recall@20 depending only on their scoring mechanism. We use mechanistic interpretability tools to explain this gap. Stylistic features such as word length, punctuation density, and function-word frequency are similarly available at every layer in every model we probe, including an off-the-shelf control encoder, suggesting that differences in feature availability do not explain the gap. Instead, causal intervention shows that the scoring mechanism determines where the encoder consolidates authorship signal. Mean pooling forces consolidation in the middle third of the network, while late interaction defers it to the upper third. We further derive this difference from the gradient structure of each scorer, and training dynamics reveal distinct learning trajectories that follow from that difference."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kulumba-etal-2026-authorship">
<titleInfo>
<title>Where Does Authorship Signal Emerge in Encoder-Based Language Models?</title>
</titleInfo>
<name type="personal">
<namePart type="given">Francis</namePart>
<namePart type="family">Kulumba</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Guillaume</namePart>
<namePart type="family">Vimont</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Laurent</namePart>
<namePart type="family">Romary</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Florian</namePart>
<namePart type="family">Cafiero</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dana</namePart>
<namePart type="family">Arad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yonatan</namePart>
<namePart type="family">Belinkov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hanjie</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nils</namePart>
<namePart type="family">Feldhus</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aaron</namePart>
<namePart type="family">Mueller</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hosein</namePart>
<namePart type="family">Mohebbi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gabriele</namePart>
<namePart type="family">Sarti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Martin</namePart>
<namePart type="family">Tutek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-476-7</identifier>
</relatedItem>
<abstract>Authorship attribution asks whether two texts were written by the same author. Models fine-tuned for this task with the same pretrained encoder, data, and loss can still differ four-fold in Recall@20 depending only on their scoring mechanism. We use mechanistic interpretability tools to explain this gap. Stylistic features such as word length, punctuation density, and function-word frequency are similarly available at every layer in every model we probe, including an off-the-shelf control encoder, suggesting that differences in feature availability do not explain the gap. Instead, causal intervention shows that the scoring mechanism determines where the encoder consolidates authorship signal. Mean pooling forces consolidation in the middle third of the network, while late interaction defers it to the upper third. We further derive this difference from the gradient structure of each scorer, and training dynamics reveal distinct learning trajectories that follow from that difference.</abstract>
<identifier type="citekey">kulumba-etal-2026-authorship</identifier>
<location>
<url>https://aclanthology.org/2026.blackboxnlp-1.47/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>818</start>
<end>834</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Where Does Authorship Signal Emerge in Encoder-Based Language Models?
%A Kulumba, Francis
%A Vimont, Guillaume
%A Romary, Laurent
%A Cafiero, Florian
%Y Arad, Dana
%Y Belinkov, Yonatan
%Y Chen, Hanjie
%Y Feldhus, Nils
%Y Mueller, Aaron
%Y Mohebbi, Hosein
%Y Sarti, Gabriele
%Y Tutek, Martin
%S Proceedings of the 9th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-476-7
%F kulumba-etal-2026-authorship
%X Authorship attribution asks whether two texts were written by the same author. Models fine-tuned for this task with the same pretrained encoder, data, and loss can still differ four-fold in Recall@20 depending only on their scoring mechanism. We use mechanistic interpretability tools to explain this gap. Stylistic features such as word length, punctuation density, and function-word frequency are similarly available at every layer in every model we probe, including an off-the-shelf control encoder, suggesting that differences in feature availability do not explain the gap. Instead, causal intervention shows that the scoring mechanism determines where the encoder consolidates authorship signal. Mean pooling forces consolidation in the middle third of the network, while late interaction defers it to the upper third. We further derive this difference from the gradient structure of each scorer, and training dynamics reveal distinct learning trajectories that follow from that difference.
%U https://aclanthology.org/2026.blackboxnlp-1.47/
%P 818-834
Markdown (Informal)
[Where Does Authorship Signal Emerge in Encoder-Based Language Models?](https://aclanthology.org/2026.blackboxnlp-1.47/) (Kulumba et al., BlackboxNLP 2026)
ACL