August 15, 2025Open Access

Evaluating scientific theories as predictive models in language neuroscience

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Question answering models improve predictability of brain responses to language stimuli, suggesting integration with qualitative theories.

QA encoding models accurately predict brain responses to language stimuli, showing improved performance over existing models.
Compact models using just 35 questions exceed baselines in predicting responses from both fMRI and ECoG data.
The approach converts qualitative theories of language into interpretable models, enhancing understanding of brain function.
Findings indicate that large language models can effectively connect qualitative scientific frameworks and quantitative data-driven models.

Singh et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

Question answering models improve predictability of brain responses to language stimuli, suggesting integration with qualitative theories.