Breaking Customized LLMs for Coding: Automated Red Teaming for Instruction Backdoor Attacks
Open Access
- 7 October 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1182-1194
Abstract
No abstract availableFunding Information
- National Key Research and Development Program of China (2024YFF0908000)
- National Natural Science Foundation of China (U24A20337)
- 111 center (B26023)
- National Research Foundation, Singapore, and the Cyber Security Agency under its National Cybersecurity R&D Programme (NCRP25-P04-TAICeN)
This publication has 20 references indexed in Scilit:
- Security of Language Models for Code: A Systematic Literature ReviewACM Transactions on Software Engineering and Methodology, 2026
- PeerGuard: Defending Multi-Agent Systems Against Backdoor Attacks Through Mutual Reasoning2025
- Large Language Model-Aware In-Context Learning for Code GenerationACM Transactions on Software Engineering and Methodology, 2025
- StagedVulBERT: Multigranular Vulnerability Detection With a Novel Pretrained Code ModelIEEE Transactions on Software Engineering, 2024
- SALLM: Security Assessment of Generated Code2024
- Poison Attack and Poison Detection on Deep Source Code Processing ModelsACM Transactions on Software Engineering and Methodology, 2024
- Stealthy Backdoor Attack for Code ModelsIEEE Transactions on Software Engineering, 2024
- Efficient Memory Management for Large Language Model Serving with PagedAttention2023
- You see what I want you to see: poisoning vulnerabilities in neural code search2022
- Backdoors in Neural Models of Source Code2022