activity
20242026
most citedA-MemGuard: A Proactive Defense Framework for LLM-Based Agent Memory

1 citations · 2 across the 6 of their papers we have counts for

collaborators

6 papers

cs.CR20261 cited

Atomicity for Agents: Exposing, Exploiting, and Mitigating TOCTOU Vulnerabilities in Browser-Use Agents

Linxi Jiang, Zhijie Liu, Haotian Luo +1

Browser-use agents are widely used for everyday tasks. They enable automated interaction with web pages through structured DOM based interfaces or vision language models operating…

cs.CR20251 cited

A-MemGuard: A Proactive Defense Framework for LLM-Based Agent Memory

Qianshan Wei, Tengchao Yang, Yaochen Wang +7

Large Language Model (LLM) agents use memory to learn from past interactions, enabling autonomous planning and decision-making in complex environments. However, this reliance on me…

cs.CR2025

AUTOVR: Automated UI Exploration for Detecting Sensitive Data Flow Exposures in Virtual Reality Apps

John Y. Kim, Chaoshun Zuo, Yanjie Zhao +1

The rise of Virtual Reality (VR) has provided developers with an unprecedented platform for creating games and applications (apps) that require distinct inputs, different from thos…

cs.CY2025

AI Safety vs. AI Security: Demystifying the Distinction and Boundaries

Zhiqiang Lin, Huan Sun, Ness Shroff

Artificial Intelligence (AI) is rapidly being integrated into critical systems across various domains, from healthcare to autonomous vehicles. While its integration brings immense…

cs.CR2025

ELFuzz: Efficient Input Generation via LLM-driven Synthesis Over Fuzzer Space

Chuyang Chen, Brendan Dolan-Gavitt, Zhiqiang Lin

Generation-based fuzzing produces appropriate test cases according to specifications of input grammars and semantic constraints to test systems and software. However, these specifi…

cs.CL2025

RedTeamCUA: Realistic Adversarial Testing of Computer-Use Agents in Hybrid Web-OS Environments

Zeyi Liao, Jaylen Jones, Linxi Jiang +5

Computer-use agents (CUAs) promise to automate complex tasks across operating systems (OS) and the web, but remain vulnerable to indirect prompt injection. Current evaluations of t…