Month: September 2023

Bridging the data gap between children and large language models

Imagine trying to learn how to bake a cake with just a tiny recipe card, while someone else has an entire library of cookbooks at their disposal. That’s the difference between children and large language models (LLMs) when it comes to language learning. LLMs have access to a massive amount of data, while children rely […]

Published on September 1, 2023

What are large language models supposed to model?

Do large language models (LLMs) offer a window into how humans process language? Contemplating the intricate relationship between artificial and biological minds is like unravelling a complex puzzle. There are multiple ways to interpret LLMs as hypotheses about humans, each with its own unique perspective. The role of (psycho)linguistic theory is crucial in this quest […]

Published on September 1, 2023

Early behavior issues have long-term societal consequences

Just like starting a car, the behavior of kindergarten students can have a lasting impact on society. A recent economic analysis has found that conduct problems among young children can lead to higher costs for society in the future. This includes expenses related to crime, medical care, and lost productivity when these individuals become adults. […]

Published on September 1, 2023