TeamOmega at SemEval-2026 Task 13: Frozen vs. Trainable Representations for Out-of-Distribution AI-Generated Code Detection: A CodeBERT Fine-Tuning Study
Year: 2026
Type: conference-paper
Abstract: This paper describes Team_Omega's submission to SemEval-2026 Task 13 (Subtask A), which focuses on detecting AI-generated code under severe cross-language and cross-domain distribution shift.We investigate whether task-specific fine-tuning improves or harms out-of-distribution (OOD) generalization by conducting a controlled comparison between two CodeBERT-based configurations: a fully frozen encoder and a partially fine-tuned model with last-layer adaptation and a deeper residual classification head.While partial fine-tuning substantially improves in-domain performance to 0.9841 macro F1 score, it leads to severe degradation under OOD evaluation, collapsing to biased predictions and achieving only 0.3026 macro F1 score.In contrast, the frozenbackbone baseline achieves lower in-domain accuracy but more stable OOD performance, reaching macro F1 score of 0.5132.Our findings highlight a critical trade-off between indomain discrimination and cross-language robustness, suggesting that preserving pretrained multilingual representations may be preferable to aggressive task adaptation in distributionshifted code detection settings.Model Venue Key Innovation Multilingual CodeBERT EMNLP'20 MLM + RTD bimodal ✓ (6 langs) UniXCoder ACL'22 Cross-modal unification ✓ (12 langs) CodeT5 ACL'21 Encoder-decoder ✓ (8 langs) CodeGen ICLR'23 Open decoder-only ✓ (multilingual)
Authors: Nahid Niyaz Shovon, Md. Naim Parvez
Institution: Rajshahi University of Engineering and Technology
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7166883377