Welcome!

I am a PhD candidate in Computer Science (AI/NLP) at Colorado State University, working with Dr. Nikhil Krishnaswamy at the SIGNAL Lab. In the summer and fall of 2024, I worked with Optum AI on their LLM-alignment team, focusing on efficient preference alignment for healthcare applications. I am currently looking for Full-Time Research Scientist/Machine Learning Engineer positions and internships.

My Research

What happens when AI begins to truly understand and align with human intentions? That’s the question I aim to answer. I specialize in developing next-generation generative AI systems (or LLMs) that are not only intelligent but also deeply aligned with human values. My work spans aligning summaries of doctor-patient interactions to designing AI “thought-partners” for enhanced collaboration, with a focus on safety, alignment, and usability. For example, most users these days have a “frictionless” experience using LLMs—but my research asks the question of whether it is possible to train and align “frictive agents”—LLMs that can cross-examine and track ToM-based beliefs in multi-user settings. I led the development of AxomiyaBERTa, the first monolingual Assamese transformer-based language model, which set new benchmarks for low-resource language processing by leveraging Assamese-specific phonological signals in transfer learning.

News

Feb’ 25: Presented my research on “Friction” Agents at DARPA’s Friction and Accountability in Conversational Transactions (FACT) PI Meeting in Stanford University!
Jan’ 25 📝 Paper on Diverse Preference Learning (DPL) in LLMs accepted at NAACL 2025 Main Conference!
🎉 Best Paper Award at Educational Data Mining (EDM) 2024
🎓 Received PhD Candidacy with Distinction, December 2024
🏆 Awarded Evolutionary Computing and AI Graduate Fellowship 2024
📝 Paper accepted at NAACL 2024 (Oral)
📝 Paper accepted at Findings of EMNLP 2024
📝 Paper accepted at LREC-COLING 2024

Selected Publications

Simultaneous Reward Distillation and Preference Learning: Get You a Language Model Who Can Do Both
Under Review
A novel approach to combining reward learning with preference optimization in language models.
DPL: Diverse Preference Learning Without A Reference Model
NAACL Main 2025
Pioneering work on preference learning that eliminates the need for reference models while maintaining diversity.
Okay, Let’s Do This! Modeling Event Coreference with Generated Rationales 🏆
NAACL 2024 (Oral)
Novel approach to event coreference using LLM-generated rationales and knowledge distillation.
Code
“Any Other Thoughts, Hedgehog?” Linking Deliberation Chains ⭐
Findings of EMNLP 2024
Proposed a novel task of linking reasoning chains in multi-agent collaborative dialogues.
Code

View all publications →

Quick Links

Contact

Email: abhijnan.nath@colostate.edu
Department of Computer Science
Colorado State University
Fort Collins, CO 80523