Hi, I'm Harsh!

I'm a final year undergraduate student at Birla Institute of Technology and Science, Pilani - Goa Campus pursuing a BE in Computer Science + MSc in Economics.

🎉 Exciting news!

• Our work on detecting implicit geographical biases in LLMs has been accepted at COLM 2025!!🎉🎉

• Our paper on building a comprehensive taxonomy for text watermarking has been accepted at NAACL 2025 Findings!🥳💥

I am interested in All Things AI! In this phase of personal growth, I believe in learning through doing! I have had the opportunity to work on several exciting projects and research initiatives with distinguished professors and mentors. I am currently working with Professor Sashank Varma (Georgia Institute of Technology) on the extent to which LLMs exhibit human-like behavioural patterns. I have worked with Professor Ali Emami (Emory University) on exploring implicit geographic bias in LLMs. I am collaborating with Grace Guo, Professor Hanspeter Pfister (Visual Computing Group, Harvard University), and Raj Sanjay Shah (Georgia Institute of Technology), to evaluate misinformation detection capabilities of VLMs for visualizations. Previously, I have worked on building a comprehensive taxonomy for text watermarking with Professor Dongwon Lee (Pennsylvania State University), and on generating free text explanations for authorship verifications with LLMs with Professor Xiang Ren (INK Lab, University of Southern California) under the mentorship of Sahana Ramnath.

In addition, I'm proficient in several programming languages, including Python, Java, and C++. I'm also familiar with several machine learning frameworks, especially Pytorch. I'm a highly motivated individual with a strong work ethic and a passion for using technology to solve complex problems.

Always happy to talk about research and have a coffee chat!