Arjun Choudhry

Presenting AQuA and JoLT at NeurIPS 2023 at New Orleans

I am an incoming ML PhD student at the School of Computational Science and Engineering at Georgia Institute of Technology, where I am advised by Prof. B Aditya Prakash at the AdityaLab.

I recently graduated with a Master’s degree at the Language Technologies Institute at Carnegie Mellon University, where I conducted research at the Auton Lab at the School of Computer Science under the guidance of Prof. Artur Dubrawski.

Over the Summer of 2025, I am working with Prof. Daphne Ippolito and Prof. Maarten Sap at the Language Technologies Institute at Carnegie Mellon University on Multimodal Stance Detection for Detection of Harmful Memes on Social Media.

My research focuses on pushing the boundaries of foundation models beyond just text and image modalities. I am particularly interested in:

Developing pragmatic multimodal foundation models capable of addressing multifaceted challenges across diverse data types including text, images, time series, audio, video, code, and tabular data.
Creating efficient inference methods for large-scale foundation models for deployment in edge devices.
Designing comprehensive evaluation frameworks and benchmarks for analyzing systemic design flaws and reasoning ability of current state-of-the-art approaches.

Through the above three research directions, my goal is to create pragmatic, efficient, multimodal models for domain-specific applications, with a focus on healthcare.

Before joining CMU as a Master’s Student in Fall 2023, I was a Visiting Research Intern (Remote) at the Auton Lab under the mentorship of Dr. Mononito Goswami and Prof. Artur Dubrawski. I worked on benchmarking efforts for label quality assessment approaches. Prior to that, I was a Research Intern Scholar at the Australian Artificial Intelligence Institute (AAII) at University of Technology Sydney and a Research Associate & Lead at the Biometric Research Laboratory at Delhi Technological University, jointly advised by Prof. Mukesh Prasad and Prof. Dinesh Kumar Vishwakarma, working on adversarial models using emotional signals to effectively detect fake news on social media. I also worked on efficient approaches for influence maximization on social networks using heuristic and meta-heuristic approaches. I was also a Research Intern at Université du Québec à Montréal under Prof. Marie-Jean Meurs, working on adversarial adaptation of language models using external unlabeled corpora for applications in Named Entity Recognition.

I earned my Bachelor’s in Information Technology from Delhi Technological University (erstwhile Delhi College of Engineering) in India, graduating top 2% in my department and top 0.5% in the university. I was advised by Prof. Dinesh Kumar Vishwakarma at the Biometric Research Laboratory, Prof. Seba Susan, and Prof. Minni Jain. I was fortunate to be selected as a MITACS Globalink Research Intern in Summer 2021 at Université du Québec à Montréal, advised by Prof. Marie-Jean Meurs, where I worked on large-scale benchmarking of multilingual language models for downstream applications in Named Entity Recognition. After my graduation, DTU subsequently awarded me with the Commendable Researcher Award in 2024.

Feel free to drop me an email at arjuncho [at] andrew [dot] cmu [dot] edu if you are interested in my research and would like to discuss more!

I am very open to collaborations and mentoring students beginning their research journey in ML and NLP. If you want to collaborate, please feel free to reach out. I would love to connect! :)

My updated CV can be found here.