2 citations · 2 across the 7 of their papers we have counts for
8 papers
Similarity Is Not Validity: Defending LLM Semantic Caches Against Poisoning
Zihan Zhang, Shuangjie Yao, Zesen Liu +8
Semantic caches reduce LLM serving costs by reusing previously generated answers for semantically similar queries. However, retrieval is based solely on embedding similarity betwee…
Tool Specifications Matter: Uncovering and Mitigating Safety Risks in AI Agents
Minghui Pan, Jiayuxuan Yang, Yuanyuan Yuan +2
AI agents extend large language models (LLMs) with external tools, enabling them to perform complex tasks and translate model outputs into consequential real-world actions. Yet LLM…
Evaluating Privilege Usage of Agents with Real-World Tools
Quan Zhang, Lianhang Fu, Lvsi Lian +5
Equipping LLM agents with real-world tools can substantially improve productivity. However, granting agents autonomy over tool use also transfers the associated privileges to both…
Emergent Risks in Generative Multi-Agent Systems
Yue Huang, Yu Jiang, Wenjie Wang +12
Multi-agent systems composed of large generative models are rapidly moving from laboratory prototypes to real-world deployments, where they jointly plan, negotiate, and allocate sh…
LSPRAG: LSP-Guided RAG for Language-Agnostic Real-Time Unit Test Generation
Gwihwan Go, Quan Zhang, Chijin Zhou +2
Automated unit test generation is essential for robust software development, yet existing approaches struggle to generalize across multiple programming languages and operate within…
Inducing Vulnerable Code Generation in LLM Coding Assistants
Binqi Zeng, Quan Zhang, Chijin Zhou +3
Due to insufficient domain knowledge, LLM coding assistants often reference related solutions from the Internet to address programming problems. However, incorporating external inf…
Human-Imperceptible Retrieval Poisoning Attacks in LLM-Powered Applications
Quan Zhang, Binqi Zeng, Chijin Zhou +3
Presently, with the assistance of advanced LLM application development frameworks, more and more LLM-powered applications can effortlessly augment the LLMs' knowledge with external…
When Fuzzing Meets LLMs: Challenges and Opportunities
Yu Jiang, Jie Liang, Fuchen Ma +8
Fuzzing, a widely-used technique for bug detection, has seen advancements through Large Language Models (LLMs). Despite their potential, LLMs face specific challenges in fuzzing. I…