Currently, I am focusing on training neural networks to extract useful and interpretable representations of text. These representation can be used for many downstream tasks such as classification, sentiment analysis, question answering, and summarization. Prior to coming to Northeastern, I was an undergraduate.
My current research interests are in the area of Machine Learning, using structured neural networks to perform tasks that humans can as well as they can. This involves both designing the structures of these architectures as well as finding methods in which to train them efficiently. I believe that the path to human-level AI lies in finding the right balance between using domain knowledge to structure these neural networks and developing ways in which to generalize them to other tasks.
As an undergraduate, I majored in Physics and Computer Science. My initial intention was to attain a PhD in physics, but after doing a couple summers of research in experimental particle physics, I became interested in Machine Learning and specifically, the idea of engineering intelligence that could one day reach and surpass the level of our own.
I would like to design a learning architecture that would allow an autonomous agent to, for example, carry on some semblance of a conversation in which the responses are not preprogrammed and the thoughts expressed are contiguous throughout the conversation.
Though neural networks have an impressive ability to solve data-intensive problems in industry today, humans do much more complicated tasks with much less data. With the right added structure, a model can perform at a similar level with much less data.
I would like to continue academic research on these topics in some setting.
I grew up in Harrison, NY a town just a half hour north of New York City by car.
I earned my undergraduate degree from Johns Hopkins University.
Currently, I am focusing on training neural networks to extract useful and interpretable representations of text. These representation can be used for many downstream tasks such as classification, sentiment analysis, question answering, and summarization. Prior to coming to Northeastern, I was an undergraduate.
My current research interests are in the area of Machine Learning, using structured neural networks to perform tasks that humans can as well as they can. This involves both designing the structures of these architectures as well as finding methods in which to train them efficiently. I believe that the path to human-level AI lies in finding the right balance between using domain knowledge to structure these neural networks and developing ways in which to generalize them to other tasks.
As an undergraduate, I majored in Physics and Computer Science. My initial intention was to attain a PhD in physics, but after doing a couple summers of research in experimental particle physics, I became interested in Machine Learning and specifically, the idea of engineering intelligence that could one day reach and surpass the level of our own.
I would like to design a learning architecture that would allow an autonomous agent to, for example, carry on some semblance of a conversation in which the responses are not preprogrammed and the thoughts expressed are contiguous throughout the conversation.
Though neural networks have an impressive ability to solve data-intensive problems in industry today, humans do much more complicated tasks with much less data. With the right added structure, a model can perform at a similar level with much less data.
I would like to continue academic research on these topics in some setting.
I grew up in Harrison, NY a town just a half hour north of New York City by car.
I earned my undergraduate degree from Johns Hopkins University.