{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T18:36:45Z","timestamp":1790879805261,"version":"4.1.0"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1013416","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2025,9,18]],"date-time":"2025-09-18T00:00:00Z","timestamp":1758153600000}}],"reference-count":27,"publisher":"Public Library of Science (PLoS)","issue":"9","license":[{"start":{"date-parts":[[2025,9,12]],"date-time":"2025-09-12T00:00:00Z","timestamp":1757635200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["IIS-1822929"],"award-info":[{"award-number":["IIS-1822929"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100015326","name":"Common Fund","doi-asserted-by":"publisher","award":["R01EY023384"],"award-info":[{"award-number":["R01EY023384"]}],"id":[{"id":"10.13039\/100015326","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100015326","name":"Common Fund","doi-asserted-by":"publisher","award":["R01EY034118"],"award-info":[{"award-number":["R01EY034118"]}],"id":[{"id":"10.13039\/100015326","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>Brain activity patterns in high-level visual cortex support accurate linear classification of visual concepts (e.g., objects or scenes). It has long been appreciated that the accuracy of linear classification in any brain area depends on the geometry of its concept manifolds\u2014sets of brain activity patterns that encode images of a concept. However, it is unclear how the geometry of concept manifolds differs between regions of visual cortex that support accurate classification and those that don\u2019t, or how it differs between visual cortex and deep neural networks (DNNs). We estimated geometric properties of concept manifolds that, per a recent theory, directly determine the accuracy of simple \u201cfew-shot\u201d linear classifiers. Using a large fMRI dataset, we show that variation in classification accuracy across human visual cortex is driven by a variation in a single geometric property: the distance between manifold centers (\u201cgeometric Signal\u201d). In contrast, variation in classification accuracy across early and mid DNN layers is driven by an increase in the effective number of manifold dimensions (\u201cDimensionality\u201d). Despite this difference in the geometric properties that affect few-shot classification performance in the brain and DNNs, we find that Signal and Dimensionality are strongly, negatively correlated: when Signal increases across brain regions or DNN layers, Dimensionality decreases, and vice versa. We conclude that early visual cortical areas and DNN layers deploy different geometric strategies for accurate linear classification of concepts, even though both are subject to the same constraint.<\/jats:p>","DOI":"10.1371\/journal.pcbi.1013416","type":"journal-article","created":{"date-parts":[[2025,9,12]],"date-time":"2025-09-12T18:01:19Z","timestamp":1757700079000},"page":"e1013416","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":3,"title":["Variation in the geometry of concept manifolds across human visual cortex"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-9760-0105","authenticated-orcid":true,"given":"Ghislain","family":"St-Yves","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6604-9155","authenticated-orcid":true,"given":"Kendrick","family":"Kay","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-4041-3157","authenticated-orcid":true,"given":"Thomas","family":"Naselaris","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2025,9,12]]},"reference":[{"issue":"23","key":"pcbi.1013416.ref001","doi-asserted-by":"crossref","first-page":"8619","DOI":"10.1073\/pnas.1403112111","article-title":"Performance-optimized hierarchical models predict neural responses in higher visual cortex","volume":"111","author":"DLK Yamins","year":"2014","journal-title":"Proc Natl Acad Sci U S A."},{"issue":"11","key":"pcbi.1013416.ref002","article-title":"Deep supervised, but not unsupervised, models may explain IT cortical representation","volume":"10","author":"S-M Khaligh-Razavi","year":"2014","journal-title":"PLoS Comput Biol."},{"issue":"11","key":"pcbi.1013416.ref003","doi-asserted-by":"crossref","first-page":"1761","DOI":"10.1038\/s41593-019-0520-2","article-title":"A deep learning framework for neuroscience","volume":"22","author":"A Richards","year":"2019","journal-title":"Nat Neurosci."},{"key":"pcbi.1013416.ref004","unstructured":"Geirhos R, Rubisch P, Michaelis C, Bethge M, Wichmann F e l i x A, Brendel W. Imagenet-trained cnns are biased towards texture; increasing shape bias improves accuracy and robustness. arXiv preprint; 2018. https:\/\/doi.org\/10.48550\/arXiv.1811.12231"},{"issue":"1","key":"pcbi.1013416.ref005","doi-asserted-by":"crossref","first-page":"1872","DOI":"10.1038\/s41467-021-22078-3","article-title":"Qualitative similarities and differences in visual object representations between brains and deep networks","volume":"12","author":"G Jacob","year":"2021","journal-title":"Nat Commun."},{"issue":"43","key":"pcbi.1013416.ref006","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2200800119","article-title":"Neural representational geometry underlies few-shot concept learning","volume":"119","author":"B Sorscher","year":"2022","journal-title":"Proc Natl Acad Sci U S A."},{"issue":"8","key":"pcbi.1013416.ref007","doi-asserted-by":"crossref","first-page":"333","DOI":"10.1016\/j.tics.2007.06.010","article-title":"Untangling invariant object recognition","volume":"11","author":"JJ DiCarlo","year":"2007","journal-title":"Trends Cogn Sci."},{"key":"pcbi.1013416.ref008","doi-asserted-by":"crossref","first-page":"113","DOI":"10.1016\/j.conb.2021.08.002","article-title":"Interpreting neural computations by examining intrinsic and embedding dimensionality of neural activity","volume":"70","author":"M Jazayeri","year":"2021","journal-title":"Curr Opin Neurobiol."},{"key":"pcbi.1013416.ref009","doi-asserted-by":"crossref","first-page":"137","DOI":"10.1016\/j.conb.2021.10.010","article-title":"Neural population geometry: An approach for understanding biological and artificial neural networks","volume":"70","author":"S Chung","year":"2021","journal-title":"Curr Opin Neurobiol."},{"issue":"27","key":"pcbi.1013416.ref010","article-title":"Representations and generalization in artificial and brain neural networks","volume":"121","author":"Q Li","year":"2024","journal-title":"Proc Natl Acad Sci U S A."},{"issue":"3","key":"pcbi.1013416.ref011","doi-asserted-by":"crossref","first-page":"413","DOI":"10.1016\/j.neuron.2020.07.040","article-title":"Integrative benchmarking to advance neurally mechanistic models of human intelligence","volume":"108","author":"M Schrimpf","year":"2020","journal-title":"Neuron."},{"issue":"28","key":"pcbi.1013416.ref012","doi-asserted-by":"crossref","DOI":"10.1126\/sciadv.abm2219","article-title":"Reassessing hierarchical correspondences between brain and deep networks through direct interface","volume":"8","author":"NJ Sexton","year":"2022","journal-title":"Sci Adv."},{"key":"pcbi.1013416.ref013","doi-asserted-by":"crossref","first-page":"580632","DOI":"10.3389\/fncom.2020.580632","article-title":"DNNBrain: A unifying toolbox for mapping deep neural networks and brains","volume":"14","author":"X Chen","year":"2020","journal-title":"Front Comput Neurosci."},{"issue":"12","key":"pcbi.1013416.ref014","doi-asserted-by":"crossref","first-page":"1090","DOI":"10.1016\/j.tics.2022.09.003","article-title":"Degrees of algorithmic equivalence between the brain and its DNN models","volume":"26","author":"PG Schyns","year":"2022","journal-title":"Trends Cogn Sci."},{"issue":"3","key":"pcbi.1013416.ref015","doi-asserted-by":"crossref","first-page":"356","DOI":"10.1038\/nn.4244","article-title":"Using goal-driven deep learning models to understand sensory cortex","volume":"19","author":"DLK Yamins","year":"2016","journal-title":"Nat Neurosci."},{"issue":"1","key":"pcbi.1013416.ref016","doi-asserted-by":"crossref","first-page":"3329","DOI":"10.1038\/s41467-023-38674-4","article-title":"Brain-optimized deep neural network models of human visual areas learn non-hierarchical representations","volume":"14","author":"G St-Yves","year":"2023","journal-title":"Nat Commun."},{"issue":"1","key":"pcbi.1013416.ref017","doi-asserted-by":"crossref","first-page":"116","DOI":"10.1038\/s41593-021-00962-x","article-title":"A massive 7T fMRI dataset to bridge cognitive neuroscience and artificial intelligence","volume":"25","author":"EJ Allen","year":"2022","journal-title":"Nat Neurosci."},{"key":"pcbi.1013416.ref018","unstructured":"Tsung-Yi Lin, Michael Maire, Serge Belongie, Lubomir Bourdev, Ross Girshick, James Hays, Pietro Perona, Deva Ramanan, C. Lawrence Zitnick, and Piotr Doll\u00e1r. Microsoft coco: Common objects in context, 2015."},{"key":"pcbi.1013416.ref019","doi-asserted-by":"crossref","unstructured":"Fellbaum C. Wordnet. Theory and applications of ontology: computer applications. Springer; 2010. p. 231\u201343.","DOI":"10.1007\/978-90-481-8847-5_10"},{"issue":"4","key":"pcbi.1013416.ref020","doi-asserted-by":"crossref","first-page":"613","DOI":"10.1038\/nn.4247","article-title":"Explicit information for category-orthogonal object properties increases along the ventral stream","volume":"19","author":"H Hong","year":"2016","journal-title":"Nat Neurosci."},{"key":"pcbi.1013416.ref021","article-title":"Intrinsic dimension of data representations in deep neural networks","volume":"32","author":"A Ansuini","year":"2019","journal-title":"Adv Neural Inform Process Syst."},{"key":"pcbi.1013416.ref022","unstructured":"Raj Magesh Gauthaman, Brice M\u00e9nard, and Michael F. Bonner. Universal scale-free representations in human visual cortex; 2024."},{"issue":"7765","key":"pcbi.1013416.ref023","doi-asserted-by":"crossref","first-page":"361","DOI":"10.1038\/s41586-019-1346-5","article-title":"High-dimensional geometry of population responses in visual cortex","volume":"571","author":"C Stringer","year":"2019","journal-title":"Nature."},{"key":"pcbi.1013416.ref024","first-page":"247","article-title":"GLMdenoise: A fast, automated technique for denoising task-based fMRI data","volume":"7","author":"KN Kay","year":"2013","journal-title":"Front Neurosci."},{"issue":"10","key":"pcbi.1013416.ref025","doi-asserted-by":"crossref","first-page":"3911","DOI":"10.1093\/cercor\/bhu277","article-title":"Probabilistic maps of visual topography in human cortex","volume":"25","author":"L Wang","year":"2015","journal-title":"Cereb Cortex."},{"issue":"2","key":"pcbi.1013416.ref026","doi-asserted-by":"crossref","first-page":"647","DOI":"10.1016\/j.neuroimage.2007.09.034","article-title":"Population receptive field estimates in human visual cortex","volume":"39","author":"SO Dumoulin","year":"2008","journal-title":"Neuroimage."},{"key":"pcbi.1013416.ref027","doi-asserted-by":"crossref","unstructured":"Deng J, Dong W, Socher R, Li L-J, Li K, Fei-Fei L. ImageNet: A large-scale hierarchical image database. In: CVPR09. 2009.","DOI":"10.1109\/CVPR.2009.5206848"}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1013416","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2025,9,18]],"date-time":"2025-09-18T00:00:00Z","timestamp":1758153600000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1013416","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,18]],"date-time":"2025-09-18T18:00:01Z","timestamp":1758218401000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1013416"}},"subtitle":[],"editor":[{"given":"Laura","family":"Dugu\u00e9","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2025,9,12]]},"references-count":27,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2025,9,12]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1013416","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2024.11.26.625280","asserted-by":"object"}]},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,12]]}}}