This work develops a lightweight knowledge-distillation framework that leverages topological features from multiple teachers to enhance model performance and robustness while avoiding the computational cost of traditional TDA. This was presented and published in ICML 2024. This work was prepared with Dr. Eunsom Jeon, Rahul Khurana and Pavan Turaga.

This work was developed with Dr. Divya Banesh and Dr. Jesus Pulido during my time as a research intern with the Data Science at Scale team in Summer '24. This was presented with >500 at the Los Alamos National Laboratory Research Symposium.

This work was developed with Dr. Pavan Turaga, Dr. Eunsom Jeon, Dr. Sinjini Mitra, Dr. Hyunglae Lee and was presented at the Fulton Forge Research Expo in Spring 2024. 

This work was prepared as part of Dynamics Summer School with Anand Iyer, Dr. Joshua Templeman, Dr. Ricardo-Mejia Alvarez, James Halverson and Kenny Decay. This paper and presentation were presented at IMAC 2026.