Hadas Raviv

I am a postdoctoral researcher at the Hasson Lab, Princeton Neuroscience Institute. I lead the computational efforts of Princeton’s 1KD team aimed at addressing fundamental questions related to human language acquisition. The project involves developing a system to collect naturalistic, multidimensional data of children’s exposure to language, as well as developing new models to explain language acquisition mechanisms.

Sixteen years ago, I fell in love with natural language processing and the question of how language semantics can be represented computationally. Since then, I have explored related challenges in both industry and academia.

In my PhD study at the Technion, Israel Institute of Technology, I developed algorithms for entity retrieval and for semantic document search under the guidance of Prof. Oren Kurland. I explored similar topics as a research intern at IBM research and Yahoo! Labs in Haifa. At AI21Labs, I worked on various NLP challenges including paraphrasing, word sense disambiguation, human-based and automatic evaluation of large language models.

The rise of large language models, their success as well as their limitations, has intrigued me to investigate human language acquisition and the relationships between data, models, machines and humans. I believe that collaborative research involving researchers from different fields associated with language and language acquisition is essential to improve our understanding of what language is.

I hold a BSc in Physics and Mathematics from the Hebrew University of Jerusalem. My Master’s degree is in Physics and I earned it at Tel Aviv university. In my Master’s thesis, I studied the effects of Diffusing Alpha-emitters Radiation Therapy (DART) on different solid tumors.

Last but not least, I am the proud mother of three children. I am passionate about education and ways that technology, specifically artificial intelligence, can benefit children of all ages. My hobbies include creating art, writing, and volunteering in our local schools.