How do we go from hearing sounds and seeing patterns to actually understanding language? And how can machines learn to do the same? In other words, how can we understand brain function, language development and intelligence through modern research in neuroscience, psychology and machine learning?
Overview
Date:
Starts 17 August 2026, 08:30Ends 17 August 2026, 17:00Location:
Language:
EnglishLast sign up date:
27 July 2026
This is the focus of this seminar at Chalmers, where three internationally leading researchers share their perspectives on how cognitive systems work – both biologically and mathematically. The seminar is organised in collaboration with the 2026 School on Analytical Connectionism.
A central idea is connectionism: that thinking emerges from interactions within networks of many simple units – neurons in the brain or artificial nodes in neural networks. These networks are used as models to study how learning and understanding arise. They are powerful, but often difficult to interpret and are therefore sometimes described as “black boxes”. New theoretical tools are now making it possible to better understand how these systems work, bringing us closer to explaining how both humans and machines learn.
Program
08:30 – Registration
08:45 – Introduction
09:00–10:30 – Paul Smolensky
10:30–11:00 – Break (coffee and refreshments)
11:00–12:30 – Caroline Rowland
12:30–14:00 – Lunch
14:00–15:30 – Michael Biehl
15:30–16:00 – Break (coffee and refreshments)
16:00–17:00 – Poster session
Speakers
Paul Smolensky (Johns Hopkins University & Microsoft Research)
A pioneer in cognitive science and neural modelling. He has developed theories that integrate symbolic reasoning with neural systems, showing how language and cognition can be understood through networks that mirror brain function. His work has had significant international impact and has also been carried out in close collaboration with Microsoft Research, where he has contributed to the theoretical foundations of advanced AI. Smolensky is a Professor at Johns Hopkins University and an internationally recognised researcher in analytical connectionism.
Caroline Rowland (Max Planck Institute for Psycholinguistics)
A leading researcher in child language development and Director of the Language Development Department at the Max Planck Institute. She studies how children acquire language through interaction and in complex environments, with a focus on the interplay between cognitive development, psychological processes and linguistic input. Her research combines empirical studies of child language acquisition with models from artificial intelligence to better understand how language develops in real-world contexts.
Michael Biehl (University of Groningen)
Professor of Machine Learning with a focus on the theoretical analysis of neural networks. His research centres on developing mathematical and statistical methods to understand how complex learning systems function. Through work on, among other things, prototype-based models, he has contributed to making AI systems more interpretable, analysable and reliable, with applications across multiple scientific domains.
Organiser
Information and Communication Technology (ICT) Area of Advance, Chalmers University of Technology, together with the 2026 School on Analytical Connectionism.
