Fourteen weeks onsite at Meta wasn’t enough for Neha Verma. Committed to finishing her deep learning project, the third-year PhD student ended up extending her internship with Meta’s Fundamental AI Research team for an additional 16 weeks.
During her internship, Verma was responsible for leading a research project, collaborating and communicating with her team members, and writing a paper explaining her work on merging transformer models—deep learning models that form the backbone of many modern natural language processing systems.
Verma says this work connected with her research at Hopkins, which she conducts under the guidance of advisors Kevin Duh, an assistant research professor of computer science, and Kenton Murray, a research scientist at the Human Language Technology Center of Excellence.
“I was working on understanding natural language processing models by looking at their underlying properties,” Verma says. “My work at Meta extended this, looking into how differently trained models can still relate to each other if we transform them to look more alike.”
Verma decided to pursue an internship at Meta to get first-hand experience in the NLP research industry.
“The field is changing so much with large language models and the real-life use of AI everywhere,” she explains. “I wanted to be able to see the world of industry research from inside one of the companies putting massive amounts of funding into developing advanced computing systems.”
Verma describes her experience at Meta as positive and rewarding. And when she wasn’t onsite at Meta HQ in Menlo Park, California, she was busy exploring the Bay Area.
“I went surfing for the first time!” she recalls. “I also went tide-pooling, to my first NFL game, and made many trips to the redwood forests. It is really so beautiful there!”
As for her future plans, Verma is waiting to see how the rapidly changing field of AI shapes up.
“It’s still changing so much, but I’m absolutely certain that my experience at Meta will help inform my eventual decision,” she says.