10.03.2026 Gemma Boleda

Since the middle of the 20th century, a fierce battle is being fought between symbolic and distributed approaches to language and cognition. In this talk, I reflect on deep learning models of language in the context of this broader discussion, and consider the possibility that LLMs represent a synthesis between the two antagonistic traditions. This possibility is supported on the known fact that deep learning architectures allow for both distributed/continuous/fuzzy and symbolic/discrete/categorical-like representations and processing; what is left to see is how modern language models make use of this flexibility. I will discuss recent research in interpretability that suggests that grammar is processed in a near-symbolic fashion in modern LLMs, while lexical semantics receives a much more distributed treatment; and I will also discuss work in progress in our group that poses some challenges to this view. I will end the talk by summarizing a recent systematic review of interpretability work on syntactic knowledge in LLMs.