Neil Zhenqiang Gong

Package Hallucination Attacks on Coding Agents through Prompt Injection in Rule Files

Secure Speculative Decoding for Large Language Models

Divide and Doubt: Diverse Distributed Poisoning for Retrieval-Augmented Generation

LongPIBench: A Long-Context Benchmark for Prompt Injection

ContextLeak: Exfiltrating LLM Agent Context via Malicious Tools

Dynamic Malicious Skills in Agentic AI

Same-Origin Policy for Agentic Browsers

Measuring Real-World Prompt Injection Attacks in LLM-based Resume Screening

When Think-with-Image Meets Safety: What Determines Multimodal Jailbreak Robustness?

CleanBase: Detecting Malicious Documents in RAG Knowledge Databases

Leave My Images Alone: Preventing Multi-Modal Large Language Models from Analyzing Images via Visual Prompt Injection

Benchmarking Vision-Language Models under Contradictory Virtual Content Attacks in Augmented Reality

A Framework for Formalizing LLM Agent Security

CREDIT: Certified Ownership Verification of Deep Neural Networks Against Model Extraction Attacks

AlignSentinel: Alignment-Aware Detection of Prompt Injection Attacks

MalTool: Malicious Tool Attacks on LLM Agents

Copyright Detective: A Forensic System to Evidence LLMs Flickering Copyright Leakage Risks

WebSentinel: Detecting and Localizing Prompt Injection Attacks for Web Agents

ObliInjection: Order-Oblivious Prompt Injection Attack to LLM Agents with Multi-source Data