TDAI Director Tanya Berger-Wolf and Member Rajiv Ramnath were co-authors on the paper, “ML Field Planner: Analyzing and Optimizing ML Pipelines for Field Research,” which earned the Best Full Paper Award and the Phil Andrews Award in the Applications & Software Track at the Practice and Experience in Advanced Research Computing (PEARC) conference.  This paper was co-developed within the National Science Foundation-funded ICICLE AI Institute, led by The Ohio State University. Their contribution was integral to the design and analysis of the project.  

Additionally, Dr. Ramnath serves as the adviser to doctoral student Ron Davies, who led this project. This advisory role guided the student’s work on AI security within high-performance computing (HPC) environments.  The poster examined how this research addresses the security risks associated with integrating Large Language Models (LLMs) like ChatGPT into powerful supercomputing interfaces. While LLMs offer an easier way for researchers to interact with complex systems, they can also expose sensitive data or waste computing resources through malicious or poorly phrased prompts.

The dual recognition at PEARC25 affirms the national impact and innovation of research emanating from The Ohio State University, the Ohio Supercomputer Center (OSC), and the ICICLE AI Institute, all key partners of TDAI.  The recognition underscores a commitment to cutting-edge AI and machine learning into reliable, safe, and efficient tools for the broader scientific community.

Read more about the research here.