CURRENT RESEARCH

Exploring Large Language Models for Secure Computer Task Automation

This research is currently being conducted under the mentorship of Dr. Carlos Rubio-Medrano at Texas A&M University-Corpus Christi.

This work has been funded by the National Science Foundation (NSF), Computing Alliance of Hispanic-Serving Institutions (CAHSI), the Texas A&M University System Louis Stokes Alliance for Minority Participation (LSAMP), University of Texas at El Paso, and Texas A&M University-Corpus Christi.

Research interests: Large Language Models (LLMs), LLM-Powered Applications (LPAs), Autonomous AI Agents, Security

OVERVIEW

This research evaluates the effectiveness of Large Language Models (LLMs) as autonomous agents within LLM-Powered Applications (LPAs), focusing on their task completion abilities and security. The study identifies key strengths, limitations, and vulnerabilities, with initial findings aimed at improving LLM-based systems for high-stakes applications. A paper is underway and to soon be submited to the Association for Computing Machinery (ACM).

MY CONTRIBUTIONS

I am responsible for designing the tasks to be tested, selecting the appropriate frameworks and tools for the research, and deploying and testing the system. Currently, I am documenting our findings in a paper to submit to the Association for Computing Machinery (ACM).

Additionally, I developed the rubric and assessment methods to ensure a thorough evaluation process aligned with the research objectives. I also created the research poster and will be presenting our findings at upcoming conferences.

Page updated

Google Sites

Report abuse