15 papers
Understanding Content Moderation in Large Language Models through Restricted Books: From Refusal to Warning
Xucheng Yu, Emily Knox, Haohan Wang
As large language models enter everyday information pipelines, understanding how they handle sensitive topics matters as much as understanding whether they handle them at all. We s…
Security--Fidelity Tradeoffs: The Hidden Cost of Prompt Injection Defense
Mitchell Hermon, Rahul Gupta, Weitong Ruan +2
We identify a security-fidelity tradeoff in defending LLMs against indirect prompt injection: defenses resist injected instructions largely by suppressing untrusted text, which cor…
Agent Primitives: Reusable Latent Building Blocks for Multi-Agent Systems
Haibo Jin, Peng Kuang, Ye Yu +2
While existing multi-agent systems (MAS) can handle complex problems by enabling collaboration among multiple agents, they are often highly task-specific, relying on manually craft…
SCI-Defense: Defending Manipulation Attacks from Generative Engine Optimization
Xucheng Yu, Haibo Jin, Huimin Zeng +1
LLM-based ranking systems are vulnerable to Generative Engine Optimization (GEO) attacks, where adversaries inject semantic signals into product descriptions to artificially boost…
TextReg: Mitigating Prompt Distributional Overfitting via Regularized Text-Space Optimization
Lucheng Fu, Ye Yu, Yiyang Wang +4
Large language models (LLMs) are highly sensitive to the prompts used to specify task objectives and behavioral constraints. Many recent prompt optimization methods iteratively rew…
Preference Tuning as Spectral Update Reorganization
Peiyan Zhang, Haibo Jin, Liying Kang +1
Preference-based post-training is usually understood through endpoint behavior, yet the learned update that produces this behavior remains largely opaque. We study RLHF and related…