Melissa Franch, PhD

mfranch.bsky.social

20 March 2026 19:59

Consistent with this, all autistic patients exhibited lower effective dimensionality in the neural subspace predictive of semantics, indicating fewer independent coding axes.

Melissa Franch, PhD

mfranch.bsky.social

20 March 2026 19:59

While single neurons in all patients showed prediction error to the most surprising words, autistic patients showed reduced population-level decoding of word surprisal, which may arise from their altered weighting of contextual information.

Melissa Franch, PhD

mfranch.bsky.social

20 March 2026 19:58

An altered tracking of past words should affect expectations of future words. We extracted word surprisal from our stories and found some of the most surprising words to be ‘bungee’, ‘sizzling’’, and ‘typewriter’, whereas expected words included ‘birthday’, ‘call’, and ‘team’.

Melissa Franch, PhD

mfranch.bsky.social

20 March 2026 19:58

Neural weighting of prior context was systematically altered in autistic patients, reflected in a broader distribution of weights. This broader distribution of prior word weighting in autism resembles the proposed weaker, broad distribution of priors in Bayesian theories for ASD.

Melissa Franch, PhD

mfranch.bsky.social

20 March 2026 19:58

For example, the meaning of ‘bat’ changes depending on whether ‘swinging’ or ‘flying’ precedes it. We computed how neural responses to a word were weighted by embeddings of preceding words that provide contextual information.