Sherin Muckatira
I am a PhD student at the University of Massachusetts, Lowell, advised by Dr. Anna Rumshisky. I am broadly interested in understanding and improving the generalization, interpretability, and efficiency of language models. My research spans training dynamics, representation learning, and the emergence of complex behavior in large models. I also explore methods for efficient pre-training to reduce computational cost while maintaining performance. Through my research, I aim to make large language model (LLM) training more transparent, efficient, and interpretable.
News
[May 2025] Completed Amazon Applied Science Internship at Alexa AI.
[May 2024] Completed an Applied Science internship at Amazon.
[April 2024] Our paper on Emergent Abilities accepted to NAACL Findings 2024!
[March 2024] Our work on ReLoRA accepted to ICLR 2024!
Education
University of Massachusetts Lowell
Lowell, Massachusetts
Ph.D. in Computer Science; GPA 4.00
September 2021 – Present
Advisor: Prof. Anna Rumshisky
Arizona State University
Tempe, Arizona
Master of Science in Electrical Engineering; GPA 3.79
August 2011 – May 2013
Sir M Visvesvaraya Institute of Technology
Bangalore, India
Bachelor of Engineering in Electronics and Communication; GPA 4.00
September 2007 – July 2011
Experience
Research Assistant
Lowell, MA
University of Massachusetts Lowell | May 2023 – Present
PI: Prof. Anna Rumshisky
- Researching ways to improve the generalization, interpretability, and efficiency of language models.
- Studying training dynamics and representation learning to better understand the emergence of complex behaviors.
- Developing and testing efficient pre-training methods that reduce computational cost without sacrificing performance.
- Working toward making LLM training more transparent and practical for both academic and applied settings.
Applied Scientist Intern
Boston, MA
Amazon | May 2025 – August 2025
PI: Dr. Rinat Khaziev
- Researched and developed multilingual evaluation methods for large language models, focusing on improving performance beyond English.
- Built and trained judge models for multi-turn conversation evaluation across multiple languages.
Applied Scientist Intern
Remote
Amazon | May 2024 – August 2024
PI: Ikkei Itoku
- Built a synthetic data generation pipeline to address challenges of data scarcity and privacy in HR analytics.
- Developed datasets of career-related documents with structured annotations.
- Fine-tuned Mistral-7B-Instruct model with synthetic data, enabling it to identify specific guidelines demonstrated by employees.
Research Aide
Tempe, AZ
Arizona State University | July 2012 – May 2013
PI: Prof. Jieping Ye
- Implemented Gene expression pattern annotation using SIFT feature extraction on images in the Berkeley Drosophila Genome Project (BDGP).
- Constructed Codebooks using Bag of Words and Sparse Coding Approach.
Senior Software Engineer
Boxborough, MA
Qualcomm | October 2016 – December 2021
- Developed firmware for the physical layer of Wireless LAN chips using the Wifi 802.11 protocol.
- Designed and implemented features such as Spectral Scan and Radar Detection.
Applications Software Engineer
Chandler, AZ
NXP | June 2013 – October 2016
- Developed signal processing applications for radio communication, focusing on transmit and receive chains on a Vector Signal Processor for Power Amplifier characterization.
- Implemented communication interfaces between host processors and co-processors to enhance functionality in Power Amplifier characterization applications.
Publications
Image-level and group-level models for Drosophila gene expression pattern annotation
BMC bioinformatics, 2013
BMC bioinformatics, 2013
Blogs
Grokking
Literature Review