Olivia and Nathalie in BabyLM

LLMs and children have different learning trajectories in determiner acquisition.

November 4-9, the Empirical Methods in Natural Languge Processing Conference hosts the third BabyLM Workshop, where undergraduate alum Olivia La Fiandra and recent Baggett Fellow Nathalie Fernandez Echeverri present "Large Language Models and Children Have Different Learning Trajectories in Determiner Acquisition," as joint first authors in collaboration with Naomi Feldman and Patrick Shafto (Rutgers). The paper is based on Olivia's honors thesis, completed in spring 2025, which Nathalie advised, as well as CHILDES data that Nathalie and Olivia annotated and analyzed the previous summer.