Large language and language-vision models show intriguing emergent behaviors, yet they receive at least three to four – and sometimes as many as six – orders of magnitude more language data than human children. What accounts for this vast difference in sample efficiency?
In this talk, Michael C. Frank will describe steps toward an ecosystem in which we can address this question. In particular, Frank will discuss the use of child language and egocentric video data for model training, along with the use of developmental data for model evaluation. This ecosystem has the potential to shed light on both the question of model efficiency as well as the nature of human learning.
Frank will be hosted by Julia Leonard (Psychology).
Michael C. Frank
Benjamin Scott Crocker Professor of Human Biology; Director, Symbolic Systems Program
Stanford University
Frank is a cognitive scientist studying children's language learning and development, with a focus on the use of large-scale datasets to understand the variability and consistency of learning across cultures.
WTI Inspiring Speaker Series
The Wu Tsai Institute presents the 2024-25 Inspiring Speaker Series.
The Institute’s new signature series features an interdisciplinary lineup of speakers studying the mind and brain from different angles, including perspectives from beyond academia. These speakers were selected for their ability to bridge and communicate across disciplines with state-of-the-art research and ideas relevant to the entire Wu Tsai Community.
Talks occur on Thursdays at 4:00 pm approximately every two weeks, beginning September 5, 2024. They take place in the Workshop on the 11th floor of the Wu Tsai Institute at 100 College Street. There will be an opportunity for questions and a reception following each talk.
All members of the Yale community are welcome. Please contact wti@yale.edu if you have any accessibility-related needs. View our upcoming events and subscribe to our monthly newsletter for updates.
09.19.24
Owen D. Jones | Vanderbilt
10.17.24
Dean Buonomano | UCLA
10.31.24
Jenn Phillips-Cremins | UPenn
11.14.24
Joel Zylberberg | York University
12.12.24
Katja Brose | Chan Zuckerberg Institute
01.09.25
Andy Leifer | Princeton
01.23.25
Peter Robin Hiesinger | Freie
02.06.25
Rich Zemel | Columbia
02.20.25
Morgan Barense | University of Toronto
03.06.25
Olaf Sporns | Indiana University
03.27.25
Karla Miller | Oxford
05.15.25
Jane Wang | Google DeepMind