2024 update: I am on the job market! If you think my experience would be a good fit for your organization or institution, reach out! You can find my CV here.
Hola amigos! I’m Navyansh Mahla, currently a pre-final year student at the Indian Institute of Technology Bombay (IIT-Bombay). My focus is on computer science and artificial intelligence. I’m currently working with Prof. Ganesh Ramakrishnan on privacy issues in large-scale machine learning models like LLMs. Recently, I’ve been involved in Federated Learning techniques for fine-tuning transformer models and Robust Adversarial Learning techniques for LLMs, in collaboration with Amazon Research India.
My research mainly explores foundational models, especially those based on transformer architectures for language and vision tasks. I’m particularly intrigued by these models’ “black box” nature and am eager to find ways to improve their logical and reasoning abilities.
In addition to my research, I have focused on privacy concerns in machine learning, especially within federated frameworks. Earlier, I worked with Dr. Kshitij Jadhav on medical image segmentation, investigating various methods for low-resource settings.
At IIT-Bombay, I’ve actively participated in extracurricular activities alongside my academic work. During my sophomore year, I was the Convener for the Web and Coding Club (WnCC), where I helped organize events, workshops, and hackathons to encourage creativity and skill development among students. I also developed the engine for CodeWars V3, India’s first bot programming contest organized by WnCC.
Currently, I am the Founding Manager of the Artificial Intelligence community (AIC) at IIT-Bombay. Our team has created InstiGPT, an educational chatbot for educational institutions. Its the first chatbot of its kind deployed ever by any university.