Model metamers reveal divergent invariances between biological and artificial neural networks
Open accessPublished version of a preprint
The DOI matches its Crossref record; 115 of the 115 references it cites were found in a registry or an index.
Abstract
Deep neural network models of sensory systems are often proposed to learn representational transformations with invariances like those in the brain. To reveal these invariances, we generated ‘model metamers’, stimuli whose activations within a model stage are matched to those of a natural stimulus. Metamers for state-of-the-art supervised and unsupervised neural network models of vision and audition were often completely unrecognizable to humans when generated from late model stages, suggesting differences between model and human invariances. Targeted model changes improved human recognizability of model metamers but did not eliminate the overall human–model discrepancy. The human recognizability of a model’s metamers was well predicted by their recognizability by other models, suggesting that models contain idiosyncratic invariances in addition to those required by the task. Metamer recognizability dissociated from both traditional brain-based benchmarks and adversarial vulnerability, revealing a distinct failure mode of existing sensory models and providing a complementary benchmark for model assessment.
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


















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
“Moreover, systematic individual differences also arise across ANNs when varying network depth, dropout, and random initializations [73,79]”
From The Umwelt Representation Hypothesis: rethinking Universality (2026) · cites this paper as [73]
Experimental — accuracy not yet measured
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
Scikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007) · Behavior Research Methods · doi:10.3758/bf03193146 · Cited by 65,971
ImageNet: 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×
“We generated metamers for multiple stages of five standard visual neural networks trained to recognize objects 26–29 (trained on the ImageNet1K dataset 30 ; Fig.”
MethodStatistical power analyses using G∗Power 3.1: tests for correlation and regression analyses
Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009) · Behavior Research Methods · doi:10.3758/brm.41.4.1149 · Cited by 37,615
PyTorch: An Imperative Style, High-Performance Deep Learning Library
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., et al. (2019) · arXiv · doi:10.48550/arxiv.1912.01703 · Cited by 31,473
Learning Transferable Visual Models From Natural Language Supervision
Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al. (2021) · arXiv · doi:10.48550/arxiv.2103.00020 · Cited by 15,293 · cited 1×
“We subsequently ran a second experiment on an additional five models pre-trained on larger datasets that became available at later stages of the project 32–34 .”
MethodAggregated Residual Transformations for Deep Neural Networks
Xie, S., Girshick, R., Dollar, P., Tu, Z., & He, K. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.634 · Cited by 12,465
Identity mappings in deep residual networks
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_38 · Cited by 10,741
Intriguing properties of neural networks
Szegedy, C., Zaremba, W., Sutskever, I., Bruna, J., Erhan, D., Goodfellow, I., & Fergus, R. (2013) · arXiv · doi:10.48550/arxiv.1312.6199 · Cited by 8,143 · cited 2×
“A known peculiarity of contemporary artificial neural networks is their vulnerability to small adversarial perturbations designed to change the class label predicted by a model 14,15 .”
BackgroundThe proof and measurement of association between two things
Spearman, C. (1904) · The American Journal of Psychology · doi:10.2307/1412159 · Cited by 8,039
Research works printed but not linked: 17
- Goodfellow, I., Lee, H., Le, Q., Saxe, A. & Ng, A. Measuring invariances in deep networks. In Advances in Neural Information Processing Systems 22 (eds Bengio, Y., Schuurmans, D., Lafferty, J., Williams, C. & Culotta, A.) 646–654 (Curran Associates, Inc., 2009).
- Geirhos, R., Temme, C. R. M. & Rauber, J. Generalisation in humans and deep neural networks. In Advances in Neural Information Processing Systems 31 (eds Bengio, S. et al.) 7538–7550 (Curran Associates, Inc., 2018).
- Zhang, R. Making convolutional networks shift-invariant again. In Proc. 36th International Conference on Machine Learning (eds Chaudhuri K., and Salakhutdinov, R.) 7324-7334 (PMLR, 2019).
- Wyszecki, G. & Stiles, W. S. Color Science 2nd edn (Wiley, 1982).
- Feather, J., Durango, A., Gonzalez, R. & McDermott, J. Metamers of neural networks reveal divergence from human perceptual systems. In Advances in Neural Information Processing Systems 32 (eds Wallach, H. et al.) 10078–10089 (Curran Associates, Inc. 2019).
- Simonyan, K. & Zisserman, A. Very deep convolutional networks for large-scale image recognition. In 3rd International Conference on Learning Representations (eds Bengio Y. & LeCun Y.) (2015)
- Krizhevsky, A., Sutskever, I. & Hinton, G. E. ImageNet classification with deep convolutional neural networks. In Advances in Neural Information Processing Systems 25 (eds Pereira, F., Burges, C. J., Bottou, L. & Weinberger, K. Q.) 1097–1105 (Curran Associates, Inc., 2012).
- Chen, T., Kornblith, S., Norouzi, M. & Hinton, G. A simple framework for contrastive learning of visual representations. In Proc. 37th International Conference on Machine Learning (eds Daumé III, H. & Singh, A.) 1597–1607 (PMLR, 2020).
- Hermann, K., Chen, T. & Kornblith, S. The origins and prevalence of texture bias in convolutional neural networks. In Advances in Neural Information Processing Systems 33 (eds Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M. F. & Lin, H.) 19000–19015 (Curran Associates, Inc., 2020).
- Madry, A., Makelov, A., Schmidt, L., Tsipras, D. & Vladu, A. Towards deep learning models resistant to adversarial attacks. In Proc. 6th International Conference on Learning Representations (eds Bengio, Y., LeCun, Y., Sainath, T., Murray, I., Ranzato, M., & Vinyals, O.) (2018).
- Goodfellow, I., Shlens, J. & Szegedy, C. Explaining and harnessing adversarial examples. In Proc. 3rd International Conference on Learning Representations (eds Bengio, Y. & LeCun, Y.) (2015).
- Hénaff, O. J. & Simoncelli, E. P. Geodesics of learned representations. In Proc. 4th International Conference on Learning Representations (eds Bengio, Y. & LeCun, Y.) (2016).
- Sabour, S., Cao, Y., Faghri, F. & Fleet, D. J. Adversarial manipulation of deep representations. In Proc. 4th International Conference on Learning Representations (eds Bengio, Y. & LeCun, Y.) (2016).
- Hendrycks, D. & Dietterich, T. Benchmarking neural network robustness to common corruptions and perturbations. In Proc. 7th International Conference on Learning Representations (eds Sainath, T., Rush, A., Levine, S. Livescu, K. & Mohamed, S.) (2019).
- Jacobsen, J.-H., Behrmann, J., Zemel, R. & Bethge, M. Excessive invariance causes adversarial vulnerability. In Proc. 7th International Conference on Learning Representations, (ICLR) (eds Sainath, T., Rush, A., Levine, S. Livescu, K. & Mohamed, S.) (2019).
- Thorat, S., Aldegheri, G. & Kietzmann, T. C. Category-orthogonal object features guide information processing in recurrent neural networks trained for object categorization. In SVRHM 2021 Workshop @ NeurIPS (2021).
- Fetaya, E., Jacobsen, J.-H., Grathwohl, W. & Zemel, R. Understanding the limitations of conditional generative models. In Proc. 8th International Conference on Learning Representations (eds Rush, A., Mohamed, S., Song, D., Cho, K., & White, M.) (2020).
Cited by
Conwell, C., Prince, J. S., Kay, K. N., Alvarez, G. A., & Konkle, T. (2022) · doi:10.1101/2022.03.28.485868 · Cited by 160
Conwell, C., Prince, J. S., Kay, K. N., Alvarez, G. A., & Konkle, T. (2024) · Nature Communications · doi:10.1038/s41467-024-53147-y · Cited by 100
Tuckute, G., Feather, J., Boebinger, D., & McDermott, J. H. (2023) · PLoS Biology · doi:10.1371/journal.pbio.3002366 · Cited by 124
Self-citationSelf-attention in vision transformers performs perceptual grouping, not attention
Mehrani, P., & Tsotsos, J. K. (2023) · Frontiers in Computer Science · doi:10.3389/fcomp.2023.1178450 · Cited by 57
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
“While these endeavors have led to notable advancements 28–31 , the critical question remains unanswered: can human-like psychological representations naturally emerge without task-specific training?”
Metadata
- Authors
- Feather, J., Leclerc, G., Mądry, A., & McDermott, J. H.
- Affiliations
- McGovern Institute for Brain Research (US)
- MIT Computer Science and Artificial Intelligence Laboratory (US)
- Harvard University (US)
- Corresponding author
- Feather, Jenelle
- McDermott, Josh H.
- Venue
- Nature Neuroscience
- Published
- 2023-10-16
- Volume
- 26
- Issue
- 11
- Pages
- 2017-2034
- ISSN
- 1097-6256 (ISSN-L)
- 1546-1726
- PMC
- PMC10620097
- OpenAlex
- W4387666878
- Cited by
- 59
- References
- 111
- Citations by year
- 2026: 26 · 2025: 19 · 2024: 9 · 2023: 3
- FWCI
- 2.71
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- Nature Neuroscience (publishedVersion, cc-by)
- Conference on Cognitive Computational Neuroscience (CCN) (publishedVersion, cc-by)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by)
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
- Funders
- U.S. Department of Health & Human Services | NIH | National Institute on Deafness and Other Communication Disorders (R01DC017970)
- National Science Foundation (BCS-1634050)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Life Sciences › Neuroscience › Cognitive Neuroscience › Visual perception and processing mechanisms
- Keywords
- invariance, deep neural network, visual models, auditory model, adversarial vulnerability
- MeSH terms
- Brain, Humans, Learning, Neural Networks, Computer
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Model metamers reveal divergent invariances between biological and artificial neural networks. https://app.cochat.ai/papers/10.1038/s41593-023-01442-0
- The paper
- Feather, J., Leclerc, G., Mądry, A., & McDermott, J. H. (2023). Model metamers reveal divergent invariances between biological and artificial neural networks. Nature Neuroscience. https://doi.org/10.1038/s41593-023-01442-0
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.