Neil Zhenqiang Gong

A2ASecBench: A Protocol-Aware Security Benchmark for Agent-to-Agent Multi-Agent Systems

EIP: Weighted Ranking of LLMs by Quantifying Question Difficulty

CREDIT: Certified Defense of Deep Neural Networks against Model Extraction Attacks

MRAG-Corrupter: Knowledge Poisoning Attacks to Multimodal Retrieval Augmented Generation

Mitigating Watermark Forgery in Generative Models via Randomized Key Selection

Forging Image Watermarks by Reversing Watermark Removal Attacks

Watermark-based Attribution of AI-Generated Content

TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models

AlignSentinel: Alignment-Aware Detection of Prompt Injection Attacks

Attacking Logic with Logic: Reasoning Injection Attack to Large Reasoning Models

Fingerprinting LLMs via Prompt Injection

PromptArmor: An Essential Baseline for Prompt Injection Defenses

SafeText: Safe Text-to-image Models via Aligning the Text Encoder

EditTrack: Detecting and Attributing AI-assisted Image Editing

WAInjectBench: Benchmarking Prompt Injection Detections for Web Agents

Competitive Advantage Attacks to Decentralized Federated Learning

A Transfer Attack to Image Watermarks

StringLLM: Understanding the String Processing Capability of Large Language Models

$DPOT_{L_0}$: Concealing Backdoored model updates in Federated Learning by Data Poisoning with $L_0$-norm-bounded Optimized Triggers

Tracing Back the Malicious Clients in Poisoning Attacks to Federated Learning

FoundationForensics: Traceback Backdoor Attacks for Vision Foundation Models

Concealing Backdoors in Federated Learning by Trigger-Optimized Data Poisoning

Watermark-based Detection and Attribution of AI-Generated Image

Stable Signature is Unstable: Removing Image Watermark from Diffusion Models

Automatically Generating Visual Hallucination Test Cases for Multimodal Large Language Models

SafeText: Safe Text-to-image Models via Aligning the Text Encoder