@inproceedings{swaminatha-rao-etal-2026-segmentation,
title = "Segmentation Fault at {S}em{E}val-2026 Task 13: A Regularization-First Approach with Generator-Based Out-of-Distribution Splits for Detecting {AI}-Generated Code",
author = "Swaminatha Rao, Lakshmi Priya and
Santhakumari Madhavan, Dhannya and
Kodeswaran, Sreya and
R, Nithila and
R, Kanmani",
editor = "Kochmar, Ekaterina and
Ghosh, Debanjan and
North, Kai and
Komachi, Mamoru",
booktitle = "Proceedings of the 20th {I}nternational {W}orkshop on {S}emantic {E}valuation (2026)",
month = jul,
year = "2026",
address = "San Diego, California, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.semeval-1.266/",
pages = "2108--2113",
ISBN = "979-8-89176-414-9",
abstract = "This paper describes our submission to SemEval-2026 Task 13 (Subtask A) on detecting AI-generated code. We fine-tune CodeBERT-base using a generator-aware out-of-distribution (OOD) validation split to better simulate unseen test generators. Strong regularization techniques, including stochastic data augmentation, dropout, weight decay, and label smoothing, are applied to prevent overfitting to generator-specific patterns. Experiments with logistic regression, UniXcoder, and vanilla CodeBERT reveal that evaluation design has a larger impact on generalization than model scale or training data volume. Our final system achieves a macro F1 score of 0.439 on the hidden test set, representing a 62{\%} relative improvement over unregularized baselines."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="swaminatha-rao-etal-2026-segmentation">
<titleInfo>
<title>Segmentation Fault at SemEval-2026 Task 13: A Regularization-First Approach with Generator-Based Out-of-Distribution Splits for Detecting AI-Generated Code</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lakshmi</namePart>
<namePart type="given">Priya</namePart>
<namePart type="family">Swaminatha Rao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dhannya</namePart>
<namePart type="family">Santhakumari Madhavan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sreya</namePart>
<namePart type="family">Kodeswaran</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nithila</namePart>
<namePart type="family">R</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kanmani</namePart>
<namePart type="family">R</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>Proceedings of the 20th International Workshop on Semantic Evaluation (2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Kochmar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Debanjan</namePart>
<namePart type="family">Ghosh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai</namePart>
<namePart type="family">North</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mamoru</namePart>
<namePart type="family">Komachi</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, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-414-9</identifier>
</relatedItem>
<abstract>This paper describes our submission to SemEval-2026 Task 13 (Subtask A) on detecting AI-generated code. We fine-tune CodeBERT-base using a generator-aware out-of-distribution (OOD) validation split to better simulate unseen test generators. Strong regularization techniques, including stochastic data augmentation, dropout, weight decay, and label smoothing, are applied to prevent overfitting to generator-specific patterns. Experiments with logistic regression, UniXcoder, and vanilla CodeBERT reveal that evaluation design has a larger impact on generalization than model scale or training data volume. Our final system achieves a macro F1 score of 0.439 on the hidden test set, representing a 62% relative improvement over unregularized baselines.</abstract>
<identifier type="citekey">swaminatha-rao-etal-2026-segmentation</identifier>
<location>
<url>https://aclanthology.org/2026.semeval-1.266/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>2108</start>
<end>2113</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Segmentation Fault at SemEval-2026 Task 13: A Regularization-First Approach with Generator-Based Out-of-Distribution Splits for Detecting AI-Generated Code
%A Swaminatha Rao, Lakshmi Priya
%A Santhakumari Madhavan, Dhannya
%A Kodeswaran, Sreya
%A R, Nithila
%A R, Kanmani
%Y Kochmar, Ekaterina
%Y Ghosh, Debanjan
%Y North, Kai
%Y Komachi, Mamoru
%S Proceedings of the 20th International Workshop on Semantic Evaluation (2026)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, USA
%@ 979-8-89176-414-9
%F swaminatha-rao-etal-2026-segmentation
%X This paper describes our submission to SemEval-2026 Task 13 (Subtask A) on detecting AI-generated code. We fine-tune CodeBERT-base using a generator-aware out-of-distribution (OOD) validation split to better simulate unseen test generators. Strong regularization techniques, including stochastic data augmentation, dropout, weight decay, and label smoothing, are applied to prevent overfitting to generator-specific patterns. Experiments with logistic regression, UniXcoder, and vanilla CodeBERT reveal that evaluation design has a larger impact on generalization than model scale or training data volume. Our final system achieves a macro F1 score of 0.439 on the hidden test set, representing a 62% relative improvement over unregularized baselines.
%U https://aclanthology.org/2026.semeval-1.266/
%P 2108-2113
Markdown (Informal)
[Segmentation Fault at SemEval-2026 Task 13: A Regularization-First Approach with Generator-Based Out-of-Distribution Splits for Detecting AI-Generated Code](https://aclanthology.org/2026.semeval-1.266/) (Swaminatha Rao et al., SemEval 2026)
ACL