Connectionists: Brenden Lake speaking on November 30 in Developing Minds global online lecture series

[Jochen Triesch]

Dear colleagues,

On November 30, the Developing Minds global online lecture series will feature Brenden Lake from New York University, USA, speaking on "Addressing two classic debates in cognitive science with deep learning“.

The live event will take place via zoom at:
10:30 am EST (Eastern Standard Time, US)
15:30 UTC (Universal Coordinated Time)
16:30 CET (Central European Time)
00:30 (Friday, December 1) JST (Japan Standard Time)

To participate please register here:
https://sites.google.com/view/developing-minds-series/home

Abstract
How can advances in machine learning best advance our understanding of human learning and development? In this talk, I'll describe two case studies using deep neural networks to address classic debates in cognitive science:

1) Can neural networks capture human-like systematic generalization? We address a 35-year-old debate started by Fodor and Pylyshyn, who argued that standard neural networks are not viable cognitive models because they lack systematic compositionality -- algebraic ability to understand and produce novel combinations from known components.  We'll show how neural network *can* achieve human-like systematic generalization when trained through meta-learning for compositionality (MLC), a new method for optimizing the compositional skills of neural networks through practice. With MLC, a neural network can match human performance and inductive biases in a head-to-head comparison of artificial language learning.

2) What ingredients do children need to learn early vocabulary words? How much is learnable from sensory input with relatively general neural networks, and how much requires stronger inductive biases (e.g., innate knowledge, domain-specific constraints, social reasoning)? Using head-mounted video recordings from a single child (61 hours of video slices over 19 months), we show how a deep neural network an acquire many word-referent mappings, generalize to novel visual referents, and achieve multi-modal alignment. These results show how critical aspects of word meaning are learnable without strong inductive biases.

Both findings emphasize the surprising power of neural network learners, even from child-scale training data, and their ability to shed new light on classical debates in cognitive science.

Selected References
   • Lake, B. M. and Baroni, M. (2023). Human-like systematic generalization through a meta-learning neural network. Nature, 623, 115-121.
   • Orhan, A. E., and Lake, B. M. (accepted-in-principle). Learning high-level visual representations from a child’s perspective without strong inductive biases. Nature Machine Intelligence.
   • Vong, W. K., Wang, W., Orhan, A. E., and Lake, B. M (under review). Grounded language acquisition through the eyes and ears of a single child. Email to request preprint.

Short Bio
Brenden M. Lake is an Assistant Professor of Psychology and Data Science at New York University. He received his M.S. and B.S. in Symbolic Systems from Stanford University in 2009, and his Ph.D. in Cognitive Science from MIT in 2014. He was a postdoctoral Data Science Fellow at NYU from 2014-2017. Brenden is a recipient of the Robert J. Glushko Prize for Outstanding Doctoral Dissertation in Cognitive Science, he is a MIT Technology Review Innovator Under 35, and his research was selected by Scientific American as one of the 10 most important advances of 2016. Brenden's research focuses on computational problems that are easier for people than they are for machines, such as learning new concepts, creating new concepts, learning-to-learn, and asking questions.

The talk will be recorded and made available for later viewing. For more information on the talk series and recordings of previous events, please visit:
https://sites.google.com/view/developing-minds-series/home 

Best regards,
 Jochen Triesch

--
Prof. Dr. Jochen Triesch
Johanna Quandt Chair for Theoretical Life Sciences
Frankfurt Institute for Advanced Studies and
Goethe University Frankfurt
http://fias.uni-frankfurt.de/~triesch/
Tel: +49 (0)69 798-47531
Fax: +49 (0)69 798-47611