Investigating Training Data Detection in AI Coders

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Tianlin, Wei, Yunxiang, Li, Zhiming, Liu, Aishan, Guo, Qing, Liu, Xianglong, Sun, Dongning, Liu, Yang
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912498211356672
author Li, Tianlin
Wei, Yunxiang
Li, Zhiming
Liu, Aishan
Guo, Qing
Liu, Xianglong
Sun, Dongning
Liu, Yang
author_facet Li, Tianlin
Wei, Yunxiang
Li, Zhiming
Liu, Aishan
Guo, Qing
Liu, Xianglong
Sun, Dongning
Liu, Yang
contents Recent advances in code large language models (CodeLLMs) have made them indispensable tools in modern software engineering. However, these models occasionally produce outputs that contain proprietary or sensitive code snippets, raising concerns about potential non-compliant use of training data, and posing risks to privacy and intellectual property. To ensure responsible and compliant deployment of CodeLLMs, training data detection (TDD) has become a critical task. While recent TDD methods have shown promise in natural language settings, their effectiveness on code data remains largely underexplored. This gap is particularly important given code's structured syntax and distinct similarity criteria compared to natural language. To address this, we conduct a comprehensive empirical study of seven state-of-the-art TDD methods on source code data, evaluating their performance across eight CodeLLMs. To support this evaluation, we introduce CodeSnitch, a function-level benchmark dataset comprising 9,000 code samples in three programming languages, each explicitly labeled as either included or excluded from CodeLLM training. Beyond evaluation on the original CodeSnitch, we design targeted mutation strategies to test the robustness of TDD methods under three distinct settings. These mutation strategies are grounded in the well-established Type-1 to Type-4 code clone detection taxonomy. Our study provides a systematic assessment of current TDD techniques for code and offers insights to guide the development of more effective and robust detection methods in the future.
format Preprint
id arxiv_https___arxiv_org_abs_2507_17389
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Investigating Training Data Detection in AI Coders
Li, Tianlin
Wei, Yunxiang
Li, Zhiming
Liu, Aishan
Guo, Qing
Liu, Xianglong
Sun, Dongning
Liu, Yang
Software Engineering
Artificial Intelligence
Recent advances in code large language models (CodeLLMs) have made them indispensable tools in modern software engineering. However, these models occasionally produce outputs that contain proprietary or sensitive code snippets, raising concerns about potential non-compliant use of training data, and posing risks to privacy and intellectual property. To ensure responsible and compliant deployment of CodeLLMs, training data detection (TDD) has become a critical task. While recent TDD methods have shown promise in natural language settings, their effectiveness on code data remains largely underexplored. This gap is particularly important given code's structured syntax and distinct similarity criteria compared to natural language. To address this, we conduct a comprehensive empirical study of seven state-of-the-art TDD methods on source code data, evaluating their performance across eight CodeLLMs. To support this evaluation, we introduce CodeSnitch, a function-level benchmark dataset comprising 9,000 code samples in three programming languages, each explicitly labeled as either included or excluded from CodeLLM training. Beyond evaluation on the original CodeSnitch, we design targeted mutation strategies to test the robustness of TDD methods under three distinct settings. These mutation strategies are grounded in the well-established Type-1 to Type-4 code clone detection taxonomy. Our study provides a systematic assessment of current TDD techniques for code and offers insights to guide the development of more effective and robust detection methods in the future.
title Investigating Training Data Detection in AI Coders
topic Software Engineering
Artificial Intelligence
url https://arxiv.org/abs/2507.17389