The DOI matches its Crossref record; 58 of the 58 references it cites were found in a registry or an index.
Brain-like functional specialization emerges spontaneously in deep neural networks
Open accessPublished version of a preprint
Abstract
The human brain contains multiple regions with distinct, often highly specialized functions, from recognizing faces to understanding language to thinking about what others are thinking. However, it remains unclear why the cortex exhibits this high degree of functional specialization in the first place. Here, we consider the case of face perception using artificial neural networks to test the hypothesis that functional segregation of face recognition in the brain reflects a computational optimization for the broader problem of visual recognition of faces and other visual categories. We find that networks trained on object recognition perform poorly on face recognition and vice versa and that networks optimized for both tasks spontaneously segregate themselves into separate systems for faces and objects. We then show functional segregation to varying degrees for other visual categories, revealing a widespread tendency for optimization (without built-in task-specific inductive biases) to lead to functional specialization in machines and, we conjecture, also brains.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures
![Fig. 1. Distinct face and object representations in singly trained CNNs while a dual-task CNN performs well. (A) Three networks with VGG16 architecture (left) were optimized, one on face identity categorization (Face CNN in red), one on object categorization (Object CNN in orange), and one on both tasks simultaneously (dual-task CNN in gray). (B) Decoding accuracy of held-out face identities and held-out object categories using activation patterns extracted from the penultimate layer [i.e., FC2 in (A)] of the Face CNN and the Object CNN. The Face CNN outperforms the Object CNN in face decoding and vice versa for object decoding. Thus, the representations optimized for each task do not naturally support the other. The dashed gray line indicates chance level (1%). Error bars indicate SEM across classification folds. (C) A dual-task CNN optimized on both tasks performed and the separate networks (% top 1 accuracy on the test set). Error bars denote 95% confidence interval (CI) bootstrapped across classes and stimuli.](https://figures.cochat.ai/paper-figures/f7fd2e615ccf47089c9fb1193e6a0236/brain-like-functional-specialization-emerges-spontaneously-in-deep-neural-figure-1.jpg)





Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Using this approach, known as neuroconnectionism, numerous studies have identified shared representational structure across brains and ANN models, as well as factors influencing this alignment [5–11]”
From The Umwelt Representation Hypothesis: rethinking Universality (2026) · cites this paper as [7]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Modular structure has been shown to spontaneously emerge in some end-to-end neural network systems in domains other than language [e.g., 203, 204]”
From Dissociating language and thought in large language models (2024) · cites this paper as [204]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 3×
“We used similar training parameters as suggested in the original VGG paper (20): stochastic gradient descent (SGD) with momentum with an initial learning rate of 10−3, a weight decay of 10−4, and momentum of 0.”
MethodImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · cited 1×
“MATERIALS AND METHODS Task-optimizing CNNs on face and object recognition To measure object categorization performance in a network trained on object categorization only (Object CNN), we trained a randomly initialized VGG16 network (20) on 423 manually sampled categories of the ILSVRC-2012 database (53).”
MethodRich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
The fusiform face area: a module in human extrastriate cortex specialized for face perception
Kanwisher, N., McDermott, J., & Chun, M. M. (1997) · The Journal of Neuroscience · doi:10.1523/jneurosci.17-11-04302.1997 · Cited by 7,738 · cited 1×
“Here, we test this third hypothesis for one of the best-established cases of functional specialization in the brain: the visual recognition of faces (5, 7, 8).”
BackgroundSelf-citationVGGFace2: A Dataset for Recognising Faces across Pose and Age
Cao, Q., Shen, L., Xie, W., Parkhi, O. M., & Zisserman, A. (2018) · 2018 13th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2018) · doi:10.1109/fg.2018.00020 · Cited by 3,032
Food-101 – Mining Discriminative Components with Random Forests
Bossard, L., Guillaumin, M., & Van Gool, L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10599-4_29 · Cited by 2,072 · cited 2×
“To test whether we would find segregation for other natural tasks, we additionally trained two dual-task networks: (i) We trained a network on object categorization and food discrimination using the Food101 ( 57 ) dataset.”
MethodPerformance-optimized hierarchical models predict neural responses in higher visual cortex
Yamins, D. L. K., Hong, H., Cadieu, C. F., Solomon, E. A., Seibert, D., & DiCarlo, J. J. (2014) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1403112111 · Cited by 1,924
Learning Face Representation from Scratch
Yi, D., Lei, Z., Liao, S., & Li, S. Z. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1411.7923 · Cited by 1,521
Understanding Neural Networks Through Deep Visualization
Yosinski, J., Clune, J., Nguyen, A. M., Fuchs, T. J., & Lipson, H. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1506.06579 · Cited by 1,358 · cited 2×
“In addition, periodic (every 10 iterations) blurring and random jittering were applied to the pixel values (50) to enhance visualization,”
BackgroundMethodDo Better ImageNet Models Transfer Better?
Kornblith, S., Shlens, J., & Le, Q. V. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00277 · Cited by 1,325 · cited 2×
“Prior work suggests that this may not necessarily be the case: CNNs optimized for generic object recognition transfer well to many other tasks (12–15), including fine-grained discrimination within a category (13).”
Background
Research works printed but not linked: 6
- S. Finger Origins Of Neuroscience: A History Of Explorations Into Brain Function (Oxford Univ. Press 2001).
- C. R. Gallistel The replacement of general-purpose learning models with adaptively specialized learning modules in The Cognitive Neurosciences M. S. Gazzaniga Ed. (MIT Press ed. 2 2000) pp. 1179–1191.
- B. Long G. Kachergis K. Agrawal M. C. Frank Detecting social information in a dense dataset of infants’ natural visual experience (PsyArXiv 2020); https://psyarxiv.com/z7tdg/.
- 10.1037/0033-295x.98.2.164 · doi:10.1037/0033-295x.98.2.164 · Paper page
- A. Nayebi N. C. L. Kong C. Zhuang J. L. Gardner A. M. Norcia D. L. K. Yamins Unsupervised models of mouse visual cortex. bioRxiv 448730 [Preprint] (2021); www.biorxiv.org/content/10.1101/2021.06.16.448730v1. · doi:10.1101/2021.06.16.448730v1
- A. Krizhevsky I. Sutskever G. E. Hinton in Advances in Neural Information Processing Systems (MIT Press 2012) pp. 1097–1105.
Cited by
Dissociating language and thought in large language models
Mahowald, K., Ivanova, A. A., Blank, I. A., Kanwisher, N., Tenenbaum, J. B., & Fedorenko, E. (2024) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2024.01.011 · Cited by 345
“Modular structure has been shown to spontaneously emerge in some end-to-end neural network systems in domains other than language (e.g., Dobs et al., 2022; Yang et al., 2019), which suggests that emergent modularity may constitute an optimal solution to many complex tasks.”
Self-citationThe neuroconnectionist research programme
Doerig, A., Sommers, R. P., Seeliger, K., Richards, B., Ismael, J., Lindsay, G. W., Kording, K. P., Konkle, T., van Gerven, M. A. J., Kriegeskorte, N., et al. (2023) · Nature Reviews Neuroscience · doi:10.1038/s41583-023-00705-w · Cited by 233
Using artificial neural networks to ask ‘why’ questions of minds and brains
Kanwisher, N., Khosla, M., & Dobs, K. (2023) · Trends in Neurosciences · doi:10.1016/j.tins.2022.12.008 · Cited by 182
“However, segregation of the highly heterogeneous visual category of food in both networks [44] and brains [72] suggests that this account is incomplete at best.”
BackgroundMethodInfluentialSelf-citationSelective responses to faces, scenes, and bodies in the ventral visual pathway of infants
Kosakowski, H. L., Cohen, M. A., Takahashi, A., Keil, B., Kanwisher, N., & Saxe, R. (2022) · Current Biology · doi:10.1016/j.cub.2021.10.064 · Cited by 163
Self-citationKhosla, M., Murty, N. A. R., & Kanwisher, N. (2022) · Current Biology · doi:10.1016/j.cub.2022.08.009 · Cited by 135
Self-citationUnderstanding Human Object Vision: A Picture Is Worth a Thousand Representations
Bracci, S., & de Beeck, H. O. (2022) · Annual Review of Psychology · doi:10.1146/annurev-psych-032720-041031 · Cited by 120
Shared functional specialization in transformer-based language models and the human brain
Kumar, S., Sumers, T. R., Yamakoshi, T., Goldstein, A., Hasson, U., Norman, K. A., Griffiths, T. L., Hawkins, R. D., & Nastase, S. A. (2024) · Nature Communications · doi:10.1038/s41467-024-49173-5 · Cited by 149
“In both language models and the human language network, emergent functional specialization likely re fl ects both architectural constraints and the statis-ticalstructureofnaturallanguage 133 – 136 .”
BackgroundA unifying framework for functional organization in early and higher ventral visual cortex
Margalit, E., Lee, H., Finzi, D., DiCarlo, J. J., Grill-Spector, K., & Yamins, D. L. K. (2024) · Neuron · doi:10.1016/j.neuron.2024.04.018 · Cited by 110
“…to categories in the ventral stream can arise under joint training for two different supervised recognition tasks, one for faces and one for objects, 102 our results demonstrate that functional specialization can be unified under a single unsupervised learning objective on a single training set.”
Blauch, N. M., Behrmann, M., & Plaut, D. C. (2022) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2112566119 · Cited by 97
“…that many of the key findings thought to support a modular view of separable, innately specified mechanisms for the recognition of different high-level domains (faces, objects, scenes) can be accounted for within a learning-based account operating under generic connectivity constraints (23, 37, 76).”
Specialized Networks for Social Cognition in the Primate Brain
Deen, B., Schwiedrzik, C. M., Sliwa, J., & Freiwald, W. A. (2023) · Annual Review of Neuroscience · doi:10.1146/annurev-neuro-102522-121410 · Cited by 70
Emerging Opportunities for Ferroelectric Field‐Effect Transistors: Integration of 2D Materials
Yang, F., Ng, H. K., Ju, X., Cai, W., Cao, J. X., Chi, D., Suwardi, A., Hu, G., Ni, Z., Wang, X. R., et al. (2024) · Advanced Functional Materials · doi:10.1002/adfm.202310438 · Cited by 61
Stoinski, L. M., Perkuhn, J., & Hebart, M. N. (2023) · Behavior Research Methods · doi:10.3758/s13428-023-02110-8 · Cited by 64
“In addition, with THINGS starting to be adopted more widely (e.g., de Varda & Strapparava, 2022; Demircan et al., 2022; Dobs et al., 2022; Frey et al., 2021; Gifford et al., 2022; Griffin, 2019; Grootswagers et al., 2022; Lam et al., 2021; Muttenthaler et al., 2022; Ratan Murty et al., 2021;…”
MethodCortical topographic motifs emerge in a self-organized map of object space
Doshi, F. R., & Konkle, T. (2023) · Science Advances · doi:10.1126/sciadv.ade8187 · Cited by 64
Modeling naturalistic face processing in humans with deep convolutional neural networks
Jiahui, G., Feilong, M., Castello, M. V. D. O., Nastase, S. A., Haxby, J. V., & Gobbini, M. I. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2304085120 · Cited by 54
Human-like object concept representations emerge naturally in multimodal large language models
Du, C., Fu, K., Wen, B., Sun, Y., Peng, J., Wei, W., Gao, Y., Wang, S., Zhang, C., Li, J., et al. (2025) · Nature Machine Intelligence · doi:10.1038/s42256-025-01049-z · Cited by 81
“Besides, we assigned intuitive labels to each dimension at the top of each row (see Methods).”
Dobs, K., Yuan, J., Martinez, J., & Kanwisher, N. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2220642120 · Cited by 48
Self-citationMeta-learned models of cognition
Binz, M., Dasgupta, I., Jagadish, A. K., Botvinick, M., Wang, J. X., & Schulz, E. (2024) · Behavioral and Brain Sciences · doi:10.1017/s0140525x23003266 · Cited by 38
Rethinking category-selectivity in human visual cortex
Ritchie, J. B., Wardle, S. G., Vaziri-Pashkam, M., Kravitz, D. J., & Baker, C. I. (2025) · Cognitive Neuroscience · doi:10.1080/17588928.2025.2543890 · Cited by 45
“● Deep neural networks (DNNs) have provided an important tool for exploring hypotheses about visual processing and cortical specialization (Blauch et al., 2022; Dobs et al., 2022; Kanwisher et al., 2023; Konkle & Alvarez, 2022).”
Wan, Q., Liu, J., Liu, T., Sun, K., & Qin, P. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.106268 · Cited by 35
The Quest for an Integrated Set of Neural Mechanisms Underlying Object Recognition in Primates
Kar, K., & DiCarlo, J. J. (2024) · Annual Review of Vision Science · doi:10.1146/annurev-vision-112823-030616 · Cited by 46
“In Section 4, we discuss how next-generation SMART models could aim to predict even finer-scale phenomena, with successes and failures yet to be determined.”
BackgroundShoham, A., Grosbard, I. D., Patashnik, O., Cohen‐Or, D., & Yovel, G. (2024) · Nature Human Behaviour · doi:10.1038/s41562-024-01816-9 · Cited by 43
“27 ,42–47).”
InfluentialA critique of motivation constructs to explain higher-order behavior: We should unpack the black box
Murayama, K., & Jach, H. K. (2024) · Behavioral and Brain Sciences · doi:10.1017/s0140525x24000025 · Cited by 40
End-to-end topographic networks as models of cortical map formation and human visual behaviour
Lu, Z., Doerig, A., Bosch, V. M., Krahmer, B., KAISER, D. H., Cichy, R. M., & Kietzmann, T. C. (2025) · Nature Human Behaviour · doi:10.1038/s41562-025-02220-7 · Cited by 41
Feng, X., Deng, N., Wen, Y., Peng, Z., Su, D., Kang, W., & Cheng, B. (2024) · ACS Nano · doi:10.1021/acsnano.4c02547 · Cited by 31
An overview of brain-like computing: Architecture, applications, and future trends
Ou, W., Xiao, S., Zhu, C., Han, W., & Zhang, Q. (2022) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2022.1041108 · Cited by 30
“Brain-like computing can help neuroscience understand the human brain more deeply and parse its structure (Amunts et al., 2016; Dobs et al., 2022).”
BackgroundNeural mechanisms of face familiarity and learning in the human amygdala and hippocampus
Cao, R., Wang, J., Brunner, P., Willie, J. T., Li, X., Rutishauser, U., Brandmeir, N. J., & Wang, S. (2023) · Cell Reports · doi:10.1016/j.celrep.2023.113520 · Cited by 29
“45 Finally, our recent simulation study systematically investigated the critical period in the development of face processing.”
Face identity coding in the deep neural network and primate brain
Wang, J., Cao, R., Brandmeir, N. J., Li, X., & Wang, S. (2022) · Communications Biology · doi:10.1038/s42003-022-03557-9 · Cited by 30
“The lesion approach can also help us better understand the functional segregation of face recognition 59 .”
BackgroundThe neural dynamics of familiar face recognition
Wiese, H., Schweinberger, S. R., & Kovács, G. (2024) · Neuroscience & Biobehavioral Reviews · doi:10.1016/j.neubiorev.2024.105943 · Cited by 25
Advancing Naturalistic Affective Science with Deep Learning
Lin, C., Bulls, L. S., Tepfer, L. J., Vyas, A. D., & Thornton, M. A. (2023) · Affective Science · doi:10.1007/s42761-023-00215-z · Cited by 22
“…have already applied deep learning to cognitive modeling, such as how information is represented in the visual cortex (Cichy & Kaiser, 2019; Dobs et al., 2022; Khaligh-Razavi & Kriegeskorte, 2014; Kohoutová et al., 2020; Konkle & Alvarez, 2022; Mehta et al., 2020; Perconti & Plebe, 2020;…”
MethodModeling Biological Face Recognition with Deep Convolutional Neural Networks
van Dyck, L. E., & Gruber, W. R. (2023) · Journal of Cognitive Neuroscience · doi:10.1162/jocn_a_02040 · Cited by 24
“, 2021), lesions studies involving pruned units (e.g., see Blauch et al., 2022; Dobs et al., 2022), or face illusions (e.”
BackgroundMethodInfluentialRepresentation Learning in Sensory Cortex: A Theory
Anselmi, F., & Poggio, T. (2022) · IEEE Access · doi:10.1109/access.2022.3208603 · Cited by 23
Chaotic recurrent neural networks for brain modelling: A review
Mattera, A., Alfieri, V., Granato, G., & Baldassarre, G. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.107079 · Cited by 24
Language-specific representation of emotion-concept knowledge causally supports emotion inference
Li, M., Su, Y.-S., Huang, H.-Y., Cheng, J., Hu, X., Zhang, X., Wang, H., Qin, Y., Wang, X., Lindquist, K. A., et al. (2024) · iScience · doi:10.1016/j.isci.2024.111401 · Cited by 29
“52,80,81 By migrating our methodology to multimodal large models, it is expected to reveal the uniqueness of different data modalities and how they should be integrated for facilitating machine emotional intelligence.”
Dynamics of specialization in neural modules under resource constraints
Béna, G., & Goodman, D. F. M. (2025) · Nature Communications · doi:10.1038/s41467-024-55188-9 · Cited by 31
“Not only that, but it has been shown that multi-task learning can drive the emergence of modularity 61,62 , and as such would be interesting toinvestigate.”
General object-based features account for letter perception
Janini, D., Hamblin, C. J., Deza, A., & Konkle, T. (2022) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1010522 · Cited by 20
Bracci, S., Mraz, J., Zeman, A. A., Leys, G., & de Beeck, H. O. (2023) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011086 · Cited by 22
“objects) leads to diverging object spaces in late DCNNs’ layers [67], which however, do not 553 generalize well to other domains [68-70].”
BackgroundCNNs reveal the computational implausibility of the expertise hypothesis
Kanwisher, N., Gupta, P., & Dobs, K. (2023) · iScience · doi:10.1016/j.isci.2023.105976 · Cited by 17
“22 Importantly, this network spontaneously developed two sets of filters in the last convolutional layer that were predominantly involved in either face or object recognition, while few fil-ters were shared between both tasks.”
BackgroundInfluentialSelf-citationComprehensive Social Trait Judgments From Faces in Autism Spectrum Disorder
Cao, R., Zhang, N., Yu, H., Webster, P. J., Paul, L. K., Li, X., Lin, C., & Wang, S. (2023) · Psychological Science · doi:10.1177/09567976231192236 · Cited by 15
Emergent Modularity in Pre-trained Transformers
Zhang, Z., Zeng, Z., Lin, Y., Xiao, C., Wang, X., Han, X., Liu, Z., Xie, R., Sun, M., & Zhou, J. (2023) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2023.findings-acl.250 · Cited by 40
Towards Explainable Deep Learning in Computational Neuroscience: Visual and Clinical Applications
Mehmood, A., Mehmood, F., & Kim, J. (2025) · Mathematics · doi:10.3390/math13203286 · Cited by 17
Yovel, G., Grosbard, I. D., & Abudarham, N. (2023) · Proceedings of the Royal Society B Biological Sciences · doi:10.1098/rspb.2023.0093 · Cited by 15
“Importantly, however, a recent study that trained a DCNN on both faces and objects, revealed that the dual-trained network spontaneously segregated to face and object units, each specialized for its own domain [64].”
BackgroundWagatsuma, N., Hidaka, A., & Tamura, H. (2022) · Frontiers in Computational Neuroscience · doi:10.3389/fncom.2022.979258 · Cited by 14
“In contrast, assuming that each channel in a layer of the AlexNet model represents selectivity to specific visual feature (Dobs et al., 2022), as shown in Figure 7, we compared the characteristics of the neuronal population activities consisting of various neurons with distinct selectivity to those of model neurons wi…”
BackgroundMethodDevelopment of Higher-Level Vision: A Network Perspective
Bourne, J. A., Cichy, R. M., Kiorpes, L., Morrone, M. C., Arcaro, M. J., & Nielsen, K. J. (2024) · Journal of Neuroscience · doi:10.1523/jneurosci.1291-24.2024 · Cited by 14
Animal models of the human brain: Successes, limitations, and alternatives
Kanwisher, N. (2025) · Current Opinion in Neurobiology · doi:10.1016/j.conb.2024.102969 · Cited by 15
“Indeed, ANNs trained only on face discrimination, which can be thought ofas the Platonic ideal of a face-specific system, can decode non faces above chance [60,81].”
Self-citationHierarchical reasoning based on perception action cycle for visual question answering
Mohamud, S. A. M., Jalali, A., & Lee, M. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.122698 · Cited by 11
A critical period for developing face recognition
Wang, J., Cao, R., Chakravarthula, P. N., Li, X., & Wang, S. (2023) · Patterns · doi:10.1016/j.patter.2023.100895 · Cited by 11
Universality of representation in biological and artificial neural networks
Hosseini, E., Casto, C., Zaslavsky, N., Conwell, C., Richardson, M., & Fedorenko, E. (2024) · doi:10.1101/2024.12.26.629294 · Cited by 10
Liu, T., Chua, Y., Ning, Y., Liu, P., Zhang, Y., Li, T., Wan, Z., Wan, Z., Chen, W., & Zhang, S. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2024.106745 · Cited by 11
Distinct visual biases affect humans and artificial intelligence in medical imaging diagnoses
McLeod, G. A., Stanley, E. A. M., Rosenal, T., & Forkert, N. D. (2025) · npj Digital Medicine · doi:10.1038/s41746-025-02226-5 · Cited by 14
“There is generally a risk that imaging features considered important by humans might be ignored by AI and vice versa 6,7 , although there are also reports of AI systems and humans each being sensitive to similar features 28,29 .”
Counting with Cilia: The Role of Morphological Computation in Basal Cognition Research
Rorot, W. (2022) · Entropy · doi:10.3390/e24111581 · Cited by 11
The state of modelling face processing in humans with deep learning
Phillips, P. J., & White, D. (2025) · British Journal of Psychology · doi:10.1111/bjop.12794 · Cited by 16
“This type of organization has been shown to emerge when DCNNs are trained to perform both object processing and face processing tasks – with different assemblies becoming specialized for face and object processing (Dobs et al., 2022).”
InfluentialChallenging the Classical View: Recognition of Identity and Expression as Integrated Processes
Schwartz, E., O’Nell, K. C., Saxe, R. R., & Anzellotti, S. (2023) · Brain Sciences · doi:10.3390/brainsci13020296 · Cited by 12
“If the DCNN models show that shared representations for scenes and faces are possible, then why does this not occur in the brain? One can speculate that there may be other mechanisms that may constrain category-specificity [38].”
BackgroundMethodFerigo, A., Iacca, G., Medvet, E., & Nadizar, G. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128811 · Cited by 9
Chen, H., Liu, B., Wang, S., Wang, X., Han, W., Wang, X., Zhu, Y., & Bi, Y. (2025) · Nature Human Behaviour · doi:10.1038/s41562-025-02357-5 · Cited by 14
“Such findings have inspired a line of model–brain fitness analyses that assessed the similarities between the performance of different models (in terms of architecture, parameters, goals, and datasets) and the activity of the VOTC to unravel the representation of the VOTC and assess the biological fitness of DNNs (Dob…”
How well do models of visual cortex generalize to out of distribution samples?
Ren, Y., & Bashivan, P. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011145 · Cited by 10
“There is recent evidence suggesting that the function performed by a single cortical column may be more complex than that performed by a single typical ANN layer [53].”
Human-like dissociations between confidence and accuracy in convolutional neural networks
Shekhar, M., & Rahnev, D. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012578 · Cited by 12
Human-like face pareidolia emerges in deep neural networks optimized for face and object recognition
Gupta, P., & Dobs, K. (2025) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012751 · Cited by 11
Self-citationDing, X., & Zhang, H. (2022) · NeuroImage · doi:10.1016/j.neuroimage.2022.119769 · Cited by 8
Han, Z., & Sereno, A. B. (2022) · Neural Computation · doi:10.1162/neco_a_01559 · Cited by 9
Tamura, H. (2024) · Scientific Reports · doi:10.1038/s41598-024-59930-7 · Cited by 8
Better artificial intelligence does not mean better models of biology
Linsley, D., Feng, P., & Serre, T. (2026) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2025.11.016 · Cited by 7
Artificial neural networks for magnetoencephalography: a review of an emerging field
Dehgan, A., Abdelhedi, H., Hadid, V., Rish, I., & Jerbi, K. (2025) · Journal of Neural Engineering · doi:10.1088/1741-2552/addd4a · Cited by 8
“Furthermore, some studies described the entire process of classification followed by similarity analysis as MVPA.”
Emergence of brain-like mirror-symmetric viewpoint tuning in convolutional neural networks
Farzmahdi, A., Zarco, W., Freiwald, W. A., Kriegeskorte, N., & Golan, T. (2024) · eLife · doi:10.7554/elife.90256 · Cited by 8
Liu, X., He, D., Zhu, M., Li, Y., Lin, L., & Cai, Q. (2024) · Developmental Cognitive Neuroscience · doi:10.1016/j.dcn.2024.101418 · Cited by 5
“Human brain development is accompanied by the maturation of functional specialization, presented as selective neural activation in certain brain regions when performing specific perceptual or cognitive tasks (Golarai et al., 2007; Dobs et al., 2022).”
BackgroundEmergence of Emotion Selectivity in Deep Neural Networks Trained to Recognize Visual Objects
Liu, P., Bo, K., Ding, M., & Fang, R. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011943 · Cited by 5
Shoham, A., Broday-Dvir, R., Yaron, I., Yovel, G., & Malach, R. (2025) · Communications Biology · doi:10.1038/s42003-025-08497-8 · Cited by 6
“PreviousstudiesthatusedDNNstopredictbrainactivitytovisual imagesprimarilyusedDNNsthatweretrainedonvisualinputs 29,32,33 .”
Unveiling the benefits of multitasking in disentangled representation formation
Feather, J., & Chung, S. (2023) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2023.05.010 · Cited by 4
Shoham, A., Broday-Dvir, R., Malach, R., & Yovel, G. (2025) · Journal of Vision · doi:10.1167/jov.25.9.1895 · Cited by 5
Sensory encoding of emotion conveyed by the face and visual context
Soderberg, K. M., Jang, G., & Kragel, P. A. (2026) · Social Cognitive and Affective Neuroscience · doi:10.1093/scan/nsag028 · Cited by 4
“2014”
A Computational Deep Learning Model of Animacy Perception in the Human Brain
Duyck, S., Costantino, A. I., Bracci, S., & de Beeck, H. O. (2022) · Communications Biology · doi:10.1038/s42003-024-07415-8 · Cited by 6
“Many recent studies have suggested that DNNs trained in image classi fi cation provide a useful computational model of human vision 13,23 – 28 .”
Amato, L. G., Vergani, A. A., Lassi, M., Carpaneto, J., Mazzeo, S., Moschini, V., Burali, R., Salvestrini, G., Fabbiani, C., Giacomucci, G., et al. (2025) · Alzheimer s Research & Therapy · doi:10.1186/s13195-025-01718-6 · Cited by 6
Drawings of THINGS: A large-scale drawing dataset of 1854 object concepts
Mukherjee, K., Huey, H., Stoinski, L. M., Hebart, M. N., Fan, J. E., & Bainbridge, W. A. (2026) · Behavior Research Methods · doi:10.3758/s13428-025-02887-w · Cited by 4
“…utilizing the THINGS concepts (Hebart et al., 2023; Grootswagers et al., 2022; Gifford et al., 2022; Kramer et al., 2023; Mukherjee et al., 2024; Dobs et al., 2022; Hebart et al., 2020; Benchetrit et al., 2023; Papale et al., 2025) facilitates comparisons of findings across different…”
Evidence Against Syntactic Encapsulation in Large Language Models
McGee, T. A., Zhang, Y., & Blank, I. (2026) · Cognitive Science · doi:10.1111/cogs.70187 · Cited by 4
A decade of comparing brains to DNNs: Progress and perspectives
Schmitt, L., Dong, D. T., Çelik, E., de Lange, F. P., & Toneva, M. (2026) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2026.06.011 · Cited by 3
Wehrheim, M. H., Alamooti, S. T., Ramezanpour, H., & Kar, K. (2026) · Nature Communications · doi:10.1038/s41467-026-76106-1 · Cited by 3
“ANN models of the ventral visual stream have shown strong alignment with primate neural responses 24 – 26 in object recognition and can partially account for neural coding in face-selective regions for identity processing 22 .”
Fast gradient-free activation maximization for neurons in spiking neural networks
Pospelov, N., Chertkov, A., Beketov, M., Oseledets, I. V., & Anokhin, K. V. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.129070 · Cited by 6
“However, modern artificial neural networks, despite their diverse architectures and training methods, exhibit similar selectivity towards complex stimuli [10].”
BackgroundUnderstanding Cortical Streams from a Computational Perspective
Han, Z., & Sereno, A. B. (2024) · Journal of Cognitive Neuroscience · doi:10.1162/jocn_a_02121 · Cited by 3
The Umwelt Representation Hypothesis: rethinking Universality
Bosch, V., Sommers, R. P., Doerig, A., & Kietzmann, T. C. (2026) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2026.07.004 · Cited by 2
Response dynamics in macaque ventral stream recapitulate the visual hierarchy
Xiao, W., Causin, T., Vinken, K., & Livingstone, M. (2026) · eNeuro · doi:10.1101/2025.11.11.686115 · Cited by 4
Category selectivity as a window into behavioral relevance
van Dyck, L. E., & Dobs, K. (2025) · Cognitive Neuroscience · doi:10.1080/17588928.2025.2590656 · Cited by 2
“In our own work, lesioning functionally distinct subnetworks in DNNs produced selective impairments in recognition performance (Dobs et al., 2022), suggesting that segregation by category is computationally advantageous.”
Self-citationIt matters who you are: biography modulates the neural dynamics of facial identity representation
Shi, Y., Gommlich, S., & Kovács, G. (2026) · Cerebral Cortex · doi:10.1093/cercor/bhag010 · Cited by 2
“In recent years, DNNs became important for measuring perceptual similarity (e.g., Schroff et al., 2015, Shoham et al., 2023) and outperformed other computational models in explaining representations (Dobs et al., 2022; Shoham et al., 2023).”
Task-structured modularity emerges in artificial networks and aligns with brain architecture
Wu, Y., Deng, S., Du, K., Mattar, M. G., Wu, Y., Bassett, D. S., Tang, H., Pan, G., & Gu, S. (2026) · Nature Machine Intelligence · doi:10.1038/s42256-026-01306-9 · Cited by 3
“RNNs are used either to replicate internal neural dynamics 27–29 and behaviour 9,30,31 , or as virtual platforms to assess whether certain brain mechanisms facilitate task learning 32,33 or emerge from the physical constraints 6,31 .”
Visual category-driven differences in memory
Steel, A., & Silson, E. H. (2022) · Routledge eBooks · doi:10.4324/9781003158134-14 · Cited by 3
Li, W., Li, J., Chu, C., Cao, D., Shi, W., Zhang, Y., & Jiang, T. (2024) · Neuroscience · doi:10.1016/j.neuroscience.2024.01.015 · Cited by 2
The face inversion effect through the lens of deep neural networks
Tousi, E., & Mur, M. (2024) · Proceedings of the Royal Society B Biological Sciences · doi:10.1098/rspb.2024.1342 · Cited by 2
Decoding face identity: A reverse-correlation approach using deep learning
Tian, X., Song, Y., & Liu, J. (2024) · Cognition · doi:10.1016/j.cognition.2024.106008 · Cited by 2
Shavlokhova, V., Vollmer, A., Stoll, C., Vollmer, M., Lang, G. M., & Saravi, B. (2024) · Symmetry · doi:10.3390/sym16020176 · Cited by 2
Shoham, A., Grosbard, I. D., Patashnik, O., Cohen‐Or, D., & Yovel, G. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.10.16.512398 · Cited by 2
Computational Mechanisms of Face Perception
Wang, J. (2023) · The Research Repository @ WVU (West Virginia University) · doi:10.33915/etd.11823 · Cited by 2
Cell-type specific population codes link inferior temporal cortex to object recognition behaviorPreprint
Muzellec, S., Sanghavi, S., & Kar, K. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.08.01.551579 · Cited by 2
The State of Modeling Face Processing in Humans with Deep LearningPreprint
White, D., & Phillips, P. J. (2024) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/hfjmq · Cited by 2
Lee, S., Ying, J., Dey, A., Jeon, Y.-N., & Issa, E. B. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.01.31.703048 · Cited by 2
Going after the bigger picture: Using high-capacity models to understand mind and brain
de Beeck, H. O., & Bracci, S. (2023) · Behavioral and Brain Sciences · doi:10.1017/s0140525x2300153x · Cited by 1
The Emergence of Topography and Hemispheric Lateralization in High-Level Vision
Behrmann, M., Blauch, N. M., & Plaut, D. C. (2025) · Quarterly Journal of Experimental Psychology · doi:10.1177/17470218251395631 · Cited by 5
“…graded lesion effects were found when units were sorted by selectivity (see Figure 3B, Blauch et al., 2022), rather than sampled spatially, confirming that the results are not explained by imperfect spatial organization, and are in-line with findings from related modeling work (Dobs et al., 2022).”
Avcu, E., Hwang, M., Brown, K., & Gow, D. W. (2023) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2023.1062230 · Cited by 1
“, 2018) to explore task-optimized feature spaces for the classification of naturalistic inputs and their implications for functional specificity in cortical processing [(Kell et al., 2018; Kell and McDermott, 2019; Dobs et al., 2022; Kanwisher et al., 2023a); see Kanwisher et al.”
BackgroundHow Can We Track Cognitive Representations with Deep Neural Networks and Intracranial EEG?
Pacheco, D. (2023) · Studies in neuroscience, psychology and behavioral economics · doi:10.1007/978-3-031-20910-9_50 · Cited by 1
Nassan, M. M. (2024) · Harvard Review of Psychiatry · doi:10.1097/hrp.0000000000000399 · Cited by 1
Jiang, T., & Zhou, G. (2024) · Cognitive Science · doi:10.1111/cogs.70020 · Cited by 1
Sleep is required for neural network plasticity in the jellyfish CassiopeaPreprint
Abrams, M. J., Zhang, L., von Emster, K., Lee, B. H., Zeigler, H., Jain, T., Jafri, A., Chen, Z., & Harland, R. M. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.05.04.538973 · Cited by 1
“Perhaps we are observing something similar to neural-network simulations that spontaneously form brain-like functional specialization when given specific tasks to achieve [51].”
BackgroundEmpirical Evidence of Consciousness in Frontier AI SystemsPreprint
Vale, M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5331919 · Cited by 1
Metadata
- Authors
- Dobs, K., Martinez, J., Kell, A. J. E., & Kanwisher, N.
- Affiliations
- McGovern Institute for Brain Research (US)
- Mortimer B. Zuckerman Mind Brain Behavior Institute (US)
- Corresponding author
- Dobs, Katharina
- Venue
- Science Advances
- Published
- 2022-03-18
- Volume
- 8
- Issue
- 11
- Pages
- eabl8913-eabl8913
- ISSN
- 2375-2548 (ISSN-L)
- PMC
- PMC8926347
- OpenAlex
- W4225899846
- Cited by
- 129
- References
- 52
- Citations by year
- 2026: 25 · 2025: 24 · 2024: 33 · 2023: 30 · 2022: 16
- FWCI
- 13.28
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by-nc
- Locations
- Science Advances (publishedVersion, cc-by-nc)
- PubMed Central (publishedVersion, cc-by-nc)
- PubMed (publishedVersion)
- Funders
- Eunice Kennedy Shriver National Institute of Child Health and Human Development (DP1HD091947)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › Face Recognition and Perception
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- functional specialization, functional segregation, face recognition (cognition), object recognition, deep neural network, computational optimization, inductive bias, brain organization, visual cortex
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.