Hello!
I recently earned an M.S. degree from Ohio State University, where I was advised by Dr. Dr. Rajiv Ramnath and Dr. Sachin Kumar, with prior guidance from Dr. D.K. Panda.
My current research focuses on understanding and enhancing reasoning capabilities in Large Language Models (LLMs), which have revolutionized natural language understanding and generation but still face critical challenges in logical deduction, systematic reasoning, and generalization. My work delves into reasoning through parametric memory, with a particular emphasis on how dataset design and model architecture influence reasoning performance. By exploring phenomena like grokking—where models unexpectedly generalize after extensive training beyond overfitting—I aim to uncover the dynamics of reasoning and generalization. I defended my thesis on this research in December 2024, presenting a comparative investigation of grokking in transformers and beyond.
I have a diverse background in research, including my role as an NLP researcher at the Wexner Medical Center, where I collaborated with Ohio’s State EMS under Dr. Rajiv Ramnath’s mentorship to explore medical reasoning and develop an AI assistant for 911 call-takers and responders. Under Dr. Brew, I led a project that utilized LongFormer’s attention mechanism to generate semantically coherent, multi-sentence responses from academic texts. As a Graduate Research Assistant at NOWLAB with Dr. D.K. Panda, I optimized NVIDIA’s Rapids, GDS, and DALI.
I have completed my Bachelor’s in Computer Science and had the opportunity to work on an NLP project with Amazon Research, categorizing product reviews, and proposed a novel graph-based summarization. Additionally, I have three years of experience developing and engineering multi-modal machine learning solutions for Cricket.com and Ninjacart.com.
For more information, you can checkout my CV here or reach out to me on my email.
News
January 2025: 🎓 Graduated with Master’s in Computer Science from Ohio State University, with a thesis in NLP.
December 2024: ⭐ Defended ~ Master’s thesis titled “Comparative Investigation of Grokking in Transformers and Beyond: A Study on Reasoning”, exploring how dataset design and model architecture influence reasoning and generalization in Large Language Models - info, slides (under review)
August 2024: Presented ~ Implicit reasoning in transformers depends on data distribution and training structure, not model size or data volume .
July 2024: Presented ~ AI assistant for 911 operators, enhancing decision accuracy with a hybrid reasoning system.
June 2024: Presented ~ Enhanced ConnectorX, reducing data migration times and optimizing large-scale AI workloads as a Research Engineer at OSU .
May 2024: Presented ~ Voicequity, identify and mitigate biases in ASRs, enhancing fairness and inclusivity - report
November 2023: Presented ~ Long-answer question generation model for contextually accurate questions from academic text - report
October 2023: Presented ~ Performance evaluation of distributed training frameworks, comparing PyTorch Distributed and Horovod for ResNet-34 and Pythia models .
June 2023: ⭐ GRA at NOWLAB
May 2023: 🎤 Talk ~ On Domain-Specific Fine-Tuning: Advancing Analytics in Medical Research at Wexner Medical Center, concluding summer research internship with the Analytics Center of Excellence at Wexner Medical Center.
May 2023: Presented ~ Investigating GPU Offloading in NVIDIA Rapids - report
April 2023: Presented ~ Performance Benchmarking of File Systems for handling massive data workloads - report
January 2023: Started my MS by thesis at Ohio State Univeristy
September 2022: 🎤 Talk ~ On Unlocking Meaning: NLP’s Role in Shaping the Future of Intelligent Systems
October 2020: Joined hdworks (cricket.com) as a full-time research engineer
September 2020: 🎓 Graduated with Bachelor’s in Computer Science
September 2020: Completed my bachelor’s thesis with Amazon Research on an NLP project focused on customer feedback
September 2020: 🎤 Talk ~ On Bridging Minds and Machines: Psychological Impact of AI Evolution
September 2020: 🎤 Talk ~ On Quantum Perspectives on Intelligence: How quantum reshapes our understanding of intelligence, new possibilities for AI
January 2017: Fouding member of Advanced Academic Center