Product
Help
Pricing
About
Jobs
Log in
Get started
Author
Add to collection
Xiaogeng Liu
Observed name:
Liu, Xiaogeng
Institutions:
Johns Hopkins University,
Tsinghua University
Observed institutions:
Johns Hopkins University,
Tsinghua University
OpenAlex ID:
A5130599059
Decomposition Attacks Across Unlinkable Identities: Limits of Stateful Defenses for LLM Services
2026
·
Sun, Bowen
, Zhao, Zhengyue
, et al.
·
arXiv (Cornell University)
PDF
KeyPooling: Measuring Where LLM API Relay Paths Collapse Prompt Cache Isolation
2026
·
Sun, Bowen
, Cai, Yixi
, et al.
·
arXiv (Cornell University)
PDF
When Claws Remember but Do Not Tell: Stealthy Memory Injection in Persistent Personal Agents
2026
·
Zhang, Yechao
, Zhao, Shiqian
, et al.
·
arXiv (Cornell University)
PDF
SafeGen-Bench: Benchmarking Safety in Image-Conditioned Text-to-Video Generation
2026
·
Ma, Yingzi
, Liu, Xiaogeng
, et al.
·
arXiv (Cornell University)
PDF
When Are Teacher Tokens Reliable? Position-Weighted On-Policy Self-Distillation for Reasoning
2026
·
Liu, Xiaogeng
, Wang, Xinyan
, et al.
·
arXiv (Cornell University)
PDF
DecodingTrust-Agent Platform (DTap): A Controllable and Interactive Red-Teaming Platform for AI Agents
2026
·
Chen, Zhaorun
, Liu, Xun
, et al.
·
arXiv (Cornell University)
PDF
Mind Your HEARTBEAT! Claw Background Execution Inherently Enables Silent Memory Pollution
2026
·
Zhang, Yechao
, Zhao, Shiqian
, et al.
·
arXiv (Cornell University)
PDF
ROM: Real-time Overthinking Mitigation via Streaming Detection and Intervention
2026
·
Wang, Xinyan
, Liu, Xiaogeng
, et al.
·
arXiv (Cornell University)
PDF
Your Cursor is Not Secure: Command Line Interface Agent Can Expose Realistic Risks Through Tactics, Techniques, and Procedures
2026
·
Luo, Weidi
, Zhang, Qiming
, et al.
·
Rare & Special e-Zone (The Hong Kong University of Science and Technology)
Show more works
View as search
Works count:
9
Citation count:
H-index:
I10-index:
Year
Topic
Type