Neural Correlates of Language Models Are Specific to Human Language

Iñigo Parra
Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models, PMLR 322:275-289, 2026.

Abstract

Previous work has shown correlations between the hidden states of large language models and fMRI brain responses, on language tasks. These correlations have been taken as evidence of the representational similarity of these models and brain states. This study tests whether these previous results are robust to several possible concerns. Specifically this study shows: (i) that the previous results are still found after dimensionality reduction, and thus are not attributable to the curse of dimensionality; (ii) that previous results are confirmed when using new measures of similarity; (iii) that correlations between brain representations and those from models are specific to models trained on human language; and (iv) that the results are dependent on the presence of positional encoding in the models. These results confirm and strengthen the results of previous research and contribute to the debate on the biological plausibility and interpretability of state-of-the-art large language models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v322-parra26a, title = {Neural Correlates of Language Models Are Specific to Human Language}, author = {Parra, I\~{n}igo}, booktitle = {Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models}, pages = {275--289}, year = {2026}, editor = {Fumero, Marco and Domine, Clementine and L"ahner, Zorah and Cannistraci, Irene and Zhao, Bo and Williams, Alex}, volume = {322}, series = {Proceedings of Machine Learning Research}, month = {06 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v322/main/assets/parra26a/parra26a.pdf}, url = {https://proceedings.mlr.press/v322/parra26a.html}, abstract = {Previous work has shown correlations between the hidden states of large language models and fMRI brain responses, on language tasks. These correlations have been taken as evidence of the representational similarity of these models and brain states. This study tests whether these previous results are robust to several possible concerns. Specifically this study shows: (i) that the previous results are still found after dimensionality reduction, and thus are not attributable to the curse of dimensionality; (ii) that previous results are confirmed when using new measures of similarity; (iii) that correlations between brain representations and those from models are specific to models trained on human language; and (iv) that the results are dependent on the presence of positional encoding in the models. These results confirm and strengthen the results of previous research and contribute to the debate on the biological plausibility and interpretability of state-of-the-art large language models.} }
Endnote
%0 Conference Paper %T Neural Correlates of Language Models Are Specific to Human Language %A Iñigo Parra %B Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models %C Proceedings of Machine Learning Research %D 2026 %E Marco Fumero %E Clementine Domine %E Zorah L"ahner %E Irene Cannistraci %E Bo Zhao %E Alex Williams %F pmlr-v322-parra26a %I PMLR %P 275--289 %U https://proceedings.mlr.press/v322/parra26a.html %V 322 %X Previous work has shown correlations between the hidden states of large language models and fMRI brain responses, on language tasks. These correlations have been taken as evidence of the representational similarity of these models and brain states. This study tests whether these previous results are robust to several possible concerns. Specifically this study shows: (i) that the previous results are still found after dimensionality reduction, and thus are not attributable to the curse of dimensionality; (ii) that previous results are confirmed when using new measures of similarity; (iii) that correlations between brain representations and those from models are specific to models trained on human language; and (iv) that the results are dependent on the presence of positional encoding in the models. These results confirm and strengthen the results of previous research and contribute to the debate on the biological plausibility and interpretability of state-of-the-art large language models.
APA
Parra, I.. (2026). Neural Correlates of Language Models Are Specific to Human Language. Proceedings of UniReps: the Third Edition of the Workshop on Unifying Representations in Neural Models, in Proceedings of Machine Learning Research 322:275-289 Available from https://proceedings.mlr.press/v322/parra26a.html.

Related Material