6 papers
Harmless Yet Harmful: Neutral Prompting Attacks for Stealthy Hallucination Steering in Agent Skills
Chia-Yi Hsu, Chia-Mu Yu, Chun-Ying Huang +1
LLM-powered coding agents increasingly participate in software development workflows by generating code, selecting dependencies, and producing package installation commands. This c…
Trust Me, Import This: Dependency Steering Attacks via Malicious Agent Skills
Yiyong Liu, Chia-Yi Hsu, Chun-Ying Huang +3
LLM-powered coding agents increasingly make software supply chain decisions. They generate imports, recommend packages, and write installation commands. Prior work showed that thes…
BADTV: Unveiling Backdoor Threats in Third-Party Task Vectors
Chia-Yi Hsu, Yu-Lin Tsai, Yu Zhe +6
Task arithmetic in large-scale pre-trained models enables agile adaptation to diverse downstream tasks without extensive retraining. By leveraging task vectors (TVs), users can per…
VP-NTK: Exploring the Benefits of Visual Prompting in Differentially Private Data Synthesis
Chia-Yi Hsu, Jia-You Chen, Yu-Lin Tsai +4
Differentially private (DP) synthetic data has become the de facto standard for releasing sensitive data. However, many DP generative models suffer from the low utility of syntheti…
Layer-Aware Task Arithmetic: Disentangling Task-Specific and Instruction-Following Knowledge
Yan-Lun Chen, Yi-Ru Wei, Chia-Yi Hsu +5
Large language models (LLMs) demonstrate strong task-specific capabilities through fine-tuning, but merging multiple fine-tuned models often leads to degraded performance due to ov…
Safe LoRA: the Silver Lining of Reducing Safety Risks when Fine-tuning Large Language Models
Chia-Yi Hsu, Yu-Lin Tsai, Chih-Hsun Lin +3
While large language models (LLMs) such as Llama-2 or GPT-4 have shown impressive zero-shot performance, fine-tuning is still necessary to enhance their performance for customized…