I'm a PhD student in Electronic Information at Fudan University, specializing in LLM security, endogenous security, and AI-enabled traditional security research. My work focuses on bridging theoretical research with practical applications in artificial intelligence and cybersecurity.
- 🔭 Current Research Focus: LLM Security, Endogenous Security, AI-enabled Traditional Security
- 🌱 Currently Exploring: Advanced techniques in kernel vulnerability detection, Large Language Models, and MCP security
- 👨💻 Affiliation: School of Computing and Intelligence Innovation, Fudan University
- 🎓 Education:
- PhD in Electronic Information, Fudan University (2025-Present)
- B.S. in Computer Science and Technology, Shanghai Maritime University
- 🏛️ Visiting Student: HKUST
- 📫 Contact: fudan@drluo.cn
- 🔗 Connect: Academic Homepage | Google Scholar
- On the Security of Tool-Invocation Prompts for LLM-Based Agentic Systems (Under Review): Empirical risk assessment of tool-invocation prompts in LLM-based agentic systems
- SoulByte: A powerful data processing tool for digital human ecosystems, transforming WeChat chat records into high-quality AI training datasets and personal knowledge bases
- NexusAI-Hub: A unified platform for managing multi-provider AI models with OpenAI-compatible APIs and real-time analytics
- Multi-Sensor Strategy Learning for Unmanned Vehicle Motion Decision: Deep reinforcement learning approach with attention mechanism and multi-sensor fusion (Published in International Journal of Intelligent Networks)
- Research on Endogenous Security Resilience Measurement: Methodology for perception-decision systems (Scientia Sinica Informationis, CCF-Top Journal)
- An Endogenous Security-Oriented Framework for Cyber Resilience Assessment: Comprehensive framework for critical infrastructures (Applied Science, SCI Compendex)
- Languages: Python, C/C++, Java, C#, SQL, Cypher
- AI/ML Frameworks: PyTorch, TensorFlow, Keras, OpenCV, YOLO, Hugging Face Transformers
- Web Technologies: Django, FastAPI, Vue.js, Vite, RESTful APIs
- Databases: MySQL, Neo4j, Redis, MongoDB
- DevOps & Tools: Git, Docker, Linux, Wireshark, Postman, ROS
- Large Language Model Security (Prompt Injection, Jailbreaking, Fine-tuning Security)
- Endogenous Security Mechanisms and Cyber Resilience
- AI-enabled Traditional Security Solutions
- Deep Reinforcement Learning and Multi-sensor Fusion
- Knowledge Graphs and Event Logic Graphs
- Network Security and Vulnerability Analysis
- 👯 Open to collaborate on: Research projects related to LLM security, endogenous security, AI-driven vulnerability detection, and innovative AI solutions
- 🤔 Looking for help with: Advancing methodologies in large language model optimization and novel approaches to AI-enabled security frameworks
- 💬 Ask me about: LLM security challenges, endogenous security architectures, AI applications in traditional security domains, or any of my research projects!
Check out my repositories to see what I'm building, and feel free to contribute or reach out! My work reflects my passion for advancing the field of AI security and creating impactful tools that bridge the gap between theoretical research and practical applications.