Sherin Muckatira

I am a PhD student at the University of Massachusetts Lowell, advised by Dr. Anna Rumshisky. My research focuses on understanding how language models learn, generalize, and acquire new capabilities during pre-training. I study training dynamics, learned representations, emergent behavior, and efficient pre-training methods. I also work on evaluating language models, especially for mathematical reasoning, creativity, and collaborative problem solving. Through my work, I aim to make language models more efficient, transparent, and measurable in settings where standard benchmarks fall short.