Critical Size Hypothesis: How Model Hyperparameters Correlate with Its Linguistic Abilities

Critical Size Hypothesis: How Model Hyperparameters Correlate with Its Linguistic Abilities Ekaterina Voloshina author Oleg Serikov author 2024-10 text Proceedings of the 2024 CLASP Conference on Multimodality and Interaction in Language Learning Amy Qiu editor Bill Noble editor David Pagmar editor Vladislav Maraev editor Nikolai Ilinykh editor Association for Computational Linguistics Gothenburg, Sweden conference publication voloshina-serikov-2024-critical https://aclanthology.org/2024.clasp-1.1/ 2024-10 1 7