Harsh Vardhan

M.Tech CSE (Research) · FLaMe.nlp Research Lab, IIIT Delhi

I'm a research master's student at IIIT Delhi, working in NLP. My thesis is on the computational analysis of propaganda and persuasion.

More recently I've been pulled toward the interpretability side of that question: what a model's internal representations encode about a construct, and how much of it survives being put into words.

Research interests

Computational propaganda, persuasion, and narrative framing
Interpretability and analysis of models: probing internal representations
Optimization methods for machine learning

Publications

Fine-grained Narrative Classification in Biased News Articles
Zeba Afroz, Harsh Vardhan, Pawan Bhakuni, Aanchal Punia, Rajdeep Kumar, Md. Shad Akhtar
LREC 2026, pp. 8431–8445 DOI arXiv

Two further papers are under review.

Education

M.Tech, Computer Science & Engineering (Research): IIIT Delhi · 2025–2027
B.E., Computer Science & Engineering (AI/ML): Chandigarh University, Mohali · 2021–2025

Experience

Teaching Assistant, IIIT Delhi

Introduction to Machine Learning (AI201)
Introduction to Programming (CS101)
Probability and Statistics (MTH201)