Zhiqiang Shen

6 papers Mohamed bin Zayed University of Artificial Intelligence

• A Frustratingly Simple Yet Highly Effective Attack Baseline: Over 90% Success Rate Against the Strong Black-box Models of GPT-4.5/4o/o1
• CoreGuard: Safeguarding Foundational Capabilities of LLMs Against Model Stealing in Edge Deployment
• FADRM: Fast and Accurate Data Residual Matching for Dataset Distillation
• Human Texts Are Outliers: Detecting LLM-generated Texts via Out-of-distribution Detection
• Open CaptchaWorld: A Comprehensive Web-based Platform for Testing and Benchmarking Multimodal LLM Agents
• Pruning Spurious Subgraphs for Graph Out-of-Distribution Generalization