mcdok at SemEval-2026 Task 13: Finetuning LLMs for Detection of Machine-Generated Code

2026-04-23

AI Summary

This work addresses the challenges of detecting machine-generated code and attributing its source in multilingual, multidomain settings. It presents the first adaptation of the mdok framework—originally designed for text generation detection—to the code domain, introducing a unified modeling approach based on fine-tuning code-aware large language models. The proposed method simultaneously supports three subtasks: binary detection of AI-generated code, attribution to specific model families, and identification of mixed or adversarial samples. Comprehensive experiments demonstrate that the system achieves competitive performance across all subtasks of SemEval-2026 Task 13, confirming the feasibility and potential for further improvement of the proposed approach.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Large Multimodal Models (LMMs)
• Multiagent Systems: Adversarial Agents

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Multi-domain detection of the machine-generated code snippets in various programming languages is a challenging task. SemEval-2026 Task~13 copes with this challenge in various angles, as a binary detection problem as well as attribution of the source. Specifically, its subtasks also cover generator LLM family detection, as well as a hybrid code co-generated by humans and machines, or adversarially modified codes hiding its origin. Our submitted systems adjusted the existing mdok approach (focused on machine-generated text detection) to these specific kinds of problems by exploring various base models, more suitable for code understanding. The results indicate that the submitted systems are competitive in all three subtasks. However, the margins from the top-performing systems are significant, and thus further improvements are possible.

Problem

• machine-generated code
• code detection
• LLM attribution
• hybrid code
• adversarial code

Innovation

• machine-generated code detection
• large language models
• code attribution
• adversarial code
• multi-domain code analysis