Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Zonghao Ying

Zonghao Ying

3 indexed papers

Recent (6 mo)
3
With code
0
Influential cites
0
Benchmarked
0

Publications per year

3
26

Top categories

Crypto×3AI×2

Frequent co-authors

Quanchen Zou3×
Deyue Zhang2×
Dongdong Yang2×
Xiangzheng Zhang2×
Wenzhuo Xu1×
Zhipeng Wei1×

Research Timeline

2026
AgentVisor: Defending LLM Agents Against Prompt Injection via Semantic Virtualization

AgentVisor is a novel defense framework that uses semantic virtualization, inspired by OS principles, to significantly reduce LLM agent vulnerability to prompt injection while maintaining high utility.

SafeHarbor: Hierarchical Memory-Augmented Guardrail for LLM Agent Safety

SafeHarbor is a novel, hierarchical memory-augmented framework that establishes context-aware decision boundaries for LLM agents, achieving state-of-the-art safety while minimizing over-refusal.

DMN: A Compositional Framework for Jailbreaking Multimodal LLMs with Multi-Image Inputs

The paper proposes DMN, a compositional jailbreak framework that utilizes distributed instructions, multimodal evidence, and a number chain task across multiple images to significantly enhance the attack success rate against multimodal LLMs.

Highlighted terms show continued research focus across papers

Papers

cs.CRcs.AIRecentMay 18, 2026

DMN: A Compositional Framework for Jailbreaking Multimodal LLMs with Multi-Image Inputs

Wenzhuo Xu, Zhipeng Wei, Zonghao Ying, Deyue Zhang +3 more

The paper proposes DMN, a compositional jailbreak framework that utilizes distributed instructions, multimodal evidence, and a number chain task across multiple images to significantly enhance the att…

View →
cs.CRcs.AIRecentMay 7, 2026

SafeHarbor: Hierarchical Memory-Augmented Guardrail for LLM Agent Safety

Zhe Liu, Zonghao Ying, Wenxin Zhang, Quanchen Zou +4 more

SafeHarbor is a novel, hierarchical memory-augmented framework that establishes context-aware decision boundaries for LLM agents, achieving state-of-the-art safety while minimizing over-refusal.

View →
cs.CRRecentApr 27, 2026

AgentVisor: Defending LLM Agents Against Prompt Injection via Semantic Virtualization

Zonghao Ying, Haozheng Wang, Jiangfan Liu, Quanchen Zou +4 more

AgentVisor is a novel defense framework that uses semantic virtualization, inspired by OS principles, to significantly reduce LLM agent vulnerability to prompt injection while maintaining high utility…

View →