logo
CanyamAI summaries for academic research
language
MobileMobile
  • Phone FeatureWeChat Mini Program
EnglishEnglish
noticeNotice
BlogBlog
delete
arrow
Return

MMAIGCD — Multimodal multilingual AI-generated code detection

delete2026-07-23
delete0
PRE
AI
H
Hamza Boulibya *
L
Laila El Ouedeghyry
Y
Youssef Azami
M
Mohamed Lamrini
H
Hamid Tairi
J
Jamal Riffi
DOI：10.1016/j.jss.2026.113038delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Novel multimodal detector combining semantic features detects AI-generated code. • Five evaluation settings: in-distribution, OOD, cross-language, paraphrasing, task-level • Robust to adversarial paraphrasing with minimal performance degradation. • Robust generalization across unseen LLMs and programming languages. • Visual and semantic modalities capture complementary patterns in source code

Journal

Journal of Systems and Software cover
Journal of Systems and Software
IF:
4.1
Papers:
5.4K
Citations:
8.4K

Organization

No organization information available