@inproceedings{lu-etal-2026-synergizing,
title = "Synergizing Stylometrics with Semantics: Dual-Path Framework for {LLM} Detection and Attribution",
author = "Lu, Xingyu and
Ma, Yumeng and
Zhou, Xiang and
Gan, Shengli and
Deng, Guiying and
Wen, Yang and
Liu, Yanbing",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.1855/",
pages = "37252--37265",
ISBN = "979-8-89176-395-1",
abstract = "The widespread application of LLMs has made MGT detection increasingly important in cyberspace security and governance. The existing detection paradigms mainly focus on statistical likelihood or deep embeddings. However, in complex applications such as short texts, derivative works, and cross-domain content, the discriminative capabilities fragility of these conventional methods increases significantly with the development of LLMs. Conversely, our research reveals that LLMs exhibit inherent style inertia. To address these limitations, this study attempts to synergize stylometrics and semantics for identifying MGT. This approach draws from the forensic perspective of experts who detect human imitation by focusing on stylistic nuances. Based on the above inspiration, we propose Stylometric-Semantic LLM Attribution (SSLA), a framework that extracts model-specific stylistic fingerprints across lexical, syntactic, and structural dimensions. SSLA employs a dual-path attention fusion architecture to dynamically integrate explicit stylistic signals with implicit semantic encodings. Extensive experiments across six LLM families demonstrate that our method achieves state-of-the-art performance. Notably, SSLA achieves a Macro-F1 score of 95.6{\%} on the challenging Wikipedia dataset, demonstrating exceptional robustness and surpassing state-of-the-art baselines like OTBDetector."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lu-etal-2026-synergizing">
<titleInfo>
<title>Synergizing Stylometrics with Semantics: Dual-Path Framework for LLM Detection and Attribution</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xingyu</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yumeng</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiang</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shengli</namePart>
<namePart type="family">Gan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Guiying</namePart>
<namePart type="family">Deng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Wen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanbing</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>The widespread application of LLMs has made MGT detection increasingly important in cyberspace security and governance. The existing detection paradigms mainly focus on statistical likelihood or deep embeddings. However, in complex applications such as short texts, derivative works, and cross-domain content, the discriminative capabilities fragility of these conventional methods increases significantly with the development of LLMs. Conversely, our research reveals that LLMs exhibit inherent style inertia. To address these limitations, this study attempts to synergize stylometrics and semantics for identifying MGT. This approach draws from the forensic perspective of experts who detect human imitation by focusing on stylistic nuances. Based on the above inspiration, we propose Stylometric-Semantic LLM Attribution (SSLA), a framework that extracts model-specific stylistic fingerprints across lexical, syntactic, and structural dimensions. SSLA employs a dual-path attention fusion architecture to dynamically integrate explicit stylistic signals with implicit semantic encodings. Extensive experiments across six LLM families demonstrate that our method achieves state-of-the-art performance. Notably, SSLA achieves a Macro-F1 score of 95.6% on the challenging Wikipedia dataset, demonstrating exceptional robustness and surpassing state-of-the-art baselines like OTBDetector.</abstract>
<identifier type="citekey">lu-etal-2026-synergizing</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.1855/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>37252</start>
<end>37265</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Synergizing Stylometrics with Semantics: Dual-Path Framework for LLM Detection and Attribution
%A Lu, Xingyu
%A Ma, Yumeng
%A Zhou, Xiang
%A Gan, Shengli
%A Deng, Guiying
%A Wen, Yang
%A Liu, Yanbing
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F lu-etal-2026-synergizing
%X The widespread application of LLMs has made MGT detection increasingly important in cyberspace security and governance. The existing detection paradigms mainly focus on statistical likelihood or deep embeddings. However, in complex applications such as short texts, derivative works, and cross-domain content, the discriminative capabilities fragility of these conventional methods increases significantly with the development of LLMs. Conversely, our research reveals that LLMs exhibit inherent style inertia. To address these limitations, this study attempts to synergize stylometrics and semantics for identifying MGT. This approach draws from the forensic perspective of experts who detect human imitation by focusing on stylistic nuances. Based on the above inspiration, we propose Stylometric-Semantic LLM Attribution (SSLA), a framework that extracts model-specific stylistic fingerprints across lexical, syntactic, and structural dimensions. SSLA employs a dual-path attention fusion architecture to dynamically integrate explicit stylistic signals with implicit semantic encodings. Extensive experiments across six LLM families demonstrate that our method achieves state-of-the-art performance. Notably, SSLA achieves a Macro-F1 score of 95.6% on the challenging Wikipedia dataset, demonstrating exceptional robustness and surpassing state-of-the-art baselines like OTBDetector.
%U https://aclanthology.org/2026.findings-acl.1855/
%P 37252-37265
Markdown (Informal)
[Synergizing Stylometrics with Semantics: Dual-Path Framework for LLM Detection and Attribution](https://aclanthology.org/2026.findings-acl.1855/) (Lu et al., Findings 2026)
ACL
- Xingyu Lu, Yumeng Ma, Xiang Zhou, Shengli Gan, Guiying Deng, Yang Wen, and Yanbing Liu. 2026. Synergizing Stylometrics with Semantics: Dual-Path Framework for LLM Detection and Attribution. In Findings of the Association for Computational Linguistics: ACL 2026, pages 37252–37265, San Diego, California, United States. Association for Computational Linguistics.