@inproceedings{shovon-parvez-2026-teamomega,
    title = "{T}eam{O}mega at {S}em{E}val-2026 Task 13: Frozen vs. Trainable Representations for Out-of-Distribution {AI}-Generated Code Detection: A {C}ode{BERT} Fine-Tuning Study",
    author = "Shovon, Nahid Niyaz  and
      Parvez, Md. Naim",
    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://preview.aclanthology.org/acl-awards/2026.semeval-1.255/",
    doi = "10.18653/v1/2026.semeval-1.255",
    pages = "2034--2039",
    ISBN = "979-8-89176-414-9",
    abstract = "We propose a CodeBERT-based system for detecting AI-generated code under severe cross-language and cross-domain distribution shift. Our approach conducts a controlled comparison between a fully frozen backbone and a partially fine-tuned configuration that unfreezes only the final transformer layer with discriminative learning rates. While partial fine-tuning substantially improves in-domain performance, the frozen backbone demonstrates stronger robustness under out-of-distribution evaluation. Our results highlight a trade-off between task adaptation and cross-language generalization in machine-generated code detection."
}