6 papers
PVDetector: Detecting Prompt Injection Attacks on Purpose-Specific LLM Agents through Policy-Violation Concept Analysis
Junhui Wang, Hangtao Zhang, Zhirun Zheng +5
Large language models (LLMs) are increasingly deployed as purpose-specific agents to handle domain-specific tasks such as customer service and code generation. These agents are exp…
The Hitchhiker's Guide to Program Analysis, Part III: Mostly Harmless LLMs
Haonan Li, Tianyang Zhou, Manu Sridharan +2
Filtering out a static-analysis warning asserts that no feasible execution can trigger the reported error, whereas confirming a bug requires only one that does. Recent work uses LL…
Defending Jailbreak Attacks on Large Language Models via Manifold Trajectory Kinetics
Hangtao Zhang, Yucheng Zhao, Sishun Liu +8
Jailbreak prompts can bypass alignment guardrails in large language models (LLMs) and elicit unsafe outputs, making reliable deployment-time detection critical. Prior detection app…
A Large-Scale Comprehensive Measurement of AI-Generated Code in Real-World Repositories
Tianhao Mao, Dongfang Zhao, Haixu Tang +2
Large language models (LLMs) are rapidly transforming software engineering by enabling developers to generate code ranging from small snippets to entire projects. As AI-assisted co…
Disassembling Obfuscated Executables with LLM
Huanyao Rong, Yue Duan, Hang Zhang +4
Disassembly is a challenging task, particularly for obfuscated executables containing junk bytes, which is designed to induce disassembly errors. Existing solutions rely on heurist…
LineBreaker: Finding Token-Inconsistency Bugs with Large Language Models
Hongbo Chen, Yifan Zhang, Xing Han +7
Token-inconsistency bugs (TIBs) involve the misuse of syntactically valid yet incorrect code tokens, such as misused variables and erroneous function invocations, which can often l…