7 papers
An Image Is Worth Ten Thousand Words: Verbose-Text Induction Attacks on VLMs
Zhi Luo, Zenghui Yuan, Wenqi Wei +2
With the remarkable success of Vision-Language Models (VLMs) on multimodal tasks, concerns regarding their deployment efficiency have become increasingly prominent. In particular,…
Can LLMs Correct Themselves? A Benchmark of Self-Correction in LLMs
Guiyao Tie, Zenghui Yuan, Zeli Zhao +11
Self-correction of large language models (LLMs) emerges as a critical component for enhancing their reasoning performance. Although various self-correction methods have been propos…
Prompt Injection Attack to Tool Selection in LLM Agents
Jiawen Shi, Zenghui Yuan, Guiyao Tie +3
Tool selection is a key component of LLM agents. A popular approach follows a two-step process - \emph{retrieval} and \emph{selection} - to pick the most appropriate tool from a to…
Optimization-based Prompt Injection Attack to LLM-as-a-Judge
Jiawen Shi, Zenghui Yuan, Yinuo Liu +4
LLM-as-a-Judge uses a large language model (LLM) to select the best response from a set of candidates for a given question. LLM-as-a-Judge has many applications such as LLM-powered…
Merge Hijacking: Backdoor Attacks to Model Merging of Large Language Models
Zenghui Yuan, Yangming Xu, Jiawen Shi +2
Model merging for Large Language Models (LLMs) directly fuses the parameters of different models finetuned on various tasks, creating a unified model for multi-domain tasks. Howeve…
BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models
Zenghui Yuan, Jiawen Shi, Pan Zhou +2
Multi-modal large language models (MLLMs) extend large language models (LLMs) to process multi-modal information, enabling them to generate responses to image-text inputs. MLLMs ha…