Number detectors spontaneously emerge in a deep neural network designed for visual object recognition
Open accessThe DOI matches its Crossref record; 45 of the 45 references it cites were found in a registry or an index.
Abstract
A deep neural network trained only on visual object recognition develops tuned number detectors reminiscent of real neurons.
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. 2 Numerosity-tuned units emerging in the HCNN. (A) Examples of the stimuli used to assess numerosity encoding. Standard stimuli contain dots of the same average radius. Dots in Area & Density stimuli have a constant total area and density across all numerosities. Dots in Shape & Convex hull stimuli have random shapes and a uniform pentagon convex hull (for numerosities >4). (B) Tuning curves for individual numerosity-selective network units. Colored curves show the average responses for each stimulus set. Black curves show the average responses over all stimulus set. Error bars indicate SE measure. PN, preferred numerosity. (C) Same as (B), but for neurons in monkey prefrontal cortex (20). Only the average responses over all stimulus sets are shown. (D) Distribution of preferred numerosities of the numerosity-selective network units. (E) Same as (D), but for real neurons recorded in monkey prefrontal cortex [data from (20)].](https://figures.cochat.ai/paper-figures/aaf715d51e0d474b8c325f99b3eb8af1/number-detectors-spontaneously-emerge-in-a-deep-neural-network-designed-for-figure-2.jpg)

![Fig. 4 Tuning properties of numerosity-selective network units. (A) Left: Average tuning curves for network units preferring each numerosity plotted on a linear scale. Right: Same tuning curves plotted on a logarithmic scale. (B) Average goodness-of-fit measure for fitting Gaussian functions to the tuning curves on different scales [Plinear-log = 0.009; Plinear-pow(1/2) = 0.003; Plinear-pow(1/3) = 0.001]. (C) SD of the best-fitting Gaussian function for each of the tuning curves of numerosity-selective network units for different scales.](https://figures.cochat.ai/paper-figures/627b66a232dc435c8d7b371f03af87de/number-detectors-spontaneously-emerge-in-a-deep-neural-network-designed-for-figure-4.jpg)
![Fig. 5 Relevance of numerosity-selective units to network performance. Average activity of numerosity-selective network units shown as a function of numerical distance between preferred numerosities and sample numerosities in the matching task. Left: Data from network units. Responses were average separately for correct trials (black) and error trials (gray). Responses during error trials are normalized to the maximum average response during correct trials (P = 0.019). Right: Same plot but for real neurons recorded from monkey prefrontal cortex [data from (20)].](https://figures.cochat.ai/paper-figures/ee8c0e1e56fb4bdbaaa7270a62d0c085/number-detectors-spontaneously-emerge-in-a-deep-neural-network-designed-for-figure-5.jpg)
![Fig. 6 Performance of the HCNN model in the numerosity matching task. (A) Left: Performance functions resulting from the discrimination of numerosities plotted on a linear scale. Each curve shows the probability of the model predicting that the sample image contains the same number of items as the test image (peak of the function). Sample numerosity is indicated above each curve. Right: Same performance functions but plotted on a logarithmic scale. (B) Average goodness-of-fit measure for fitting Gaussian functions to the performance tuning curves on different scales [Plinear-log = 0.003; Plinear-pow(1/2) = 0.049; Plinear-pow(1/3) = 0.016]. *P < 0.05. (C) SD of the best-fitting Gaussian function for each of the performance tuning curves for different scales.](https://figures.cochat.ai/paper-figures/f8bf766d1b534ddcaa16bedda91e1a87/number-detectors-spontaneously-emerge-in-a-deep-neural-network-designed-for-figure-6.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
“Models trained to process images, for example, spontaneously learned how to count (30, 31) and differentially process central and peripheral image areas (32), as well as experience human-like optical illusions (33).”
From Evaluating large language models in theory of mind tasks (2024) · cites this paper as [30]
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
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 1×
“Here, we use a hierarchical convolutional neural network (HCNN), a class of biologically inspired models that have recently achieved great success in computer vision applications (14, 15) and in the modeling of the ventral visual stream (16, 17).”
MethodImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 4×
“Network units that had the same receptive fields in convolutional layers competed with each other using a simple form of lateral inhibition (14).”
BackgroundMethodInfluentialImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380
(2014) · Cited by 33,419
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818 · cited 1×
“Before applying the activation function, the outputs of the convolution operation were normalized to have zero mean and unit SD [batch normalization; (40)].”
Method(2010) · Cited by 14,134
Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex
Hubel, D. H., & Wiesel, T. N. (1962) · The Journal of Physiology · doi:10.1113/jphysiol.1962.sp006837 · Cited by 13,680 · cited 1×
“The network model was an instance of HCNNs (18), originally inspired by the discovery of simple and complex cells in early visual cortex (19).”
Method(2017) · Cited by 11,625
An overview of gradient descent optimization algorithms
Ruder, S. (2016) · arXiv · doi:10.48550/arxiv.1609.04747 · Cited by 4,962 · cited 1×
“The minimization was performed using mini-batch gradient descent (43) with a batch size of 256 images, a learning rate of 0.”
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
Research works printed but not linked: 3
- S. Dehaene The Number Sense (Oxford Univ. Press 1999).
- Y. LeCun Y. Bengio Convolutional networks for images speech and time-series in The Handbook of Brain Theory and Neural Networks (MIT Press 1995) vol. 3361 pp. 255–258.
- S. Carey The Origin of Concepts: Oxford Series in Cognitive Development (Oxford Univ. Press 2009).
Cited by
Evaluating large language models in theory of mind tasks
Kosinski, M. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2405460121 · Cited by 339
“Models trained to process images, for example, spontaneously learned how to count ( 30 , 31 ) and differentially process central and peripheral image areas ( 32 ), as well as experience human - like optical illusions ( 33 ).”
BackgroundUsing 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
“Analyses in this vein have revealed the spontaneous emergence of face-selective [88,89] and number-selective units [90,91] in ANNs.”
BackgroundThe Adaptive Value of Numerical Competence
Nieder, A. (2020) · Trends in Ecology & Evolution · doi:10.1016/j.tree.2020.02.009 · Cited by 162
“In support of a putatively innate ANS, artificial number neurons obeying Weber’s and Fechner’s law spontaneously emerged also in a deep network that was merely trained on visual object recognition, never on numbers [133].”
BackgroundSelf-citationBrain-like functional specialization emerges spontaneously in deep neural networks
Dobs, K., Martinez, J., Kell, A. J. E., & Kanwisher, N. (2022) · Science Advances · doi:10.1126/sciadv.abl8913 · Cited by 129
“…then the same should be true for any computational system, including CNNs. 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 ).”
BackgroundEvaluating Large Language Models in Theory of Mind Tasks
Kosinski, M. (2023) · arXiv · doi:10.48550/arxiv.2302.02083 · Cited by 339
“Models trained to process images, for example, spontaneously learned how to count ( 30 , 31 ) and differentially process central and peripheral image areas ( 32 ), as well as experience human - like optical illusions ( 33 ).”
Background
Metadata
- Authors
- Nasr, K., Viswanathan, P., & Nieder, A.
- Affiliations
- University of Tübingen (DE)
- Corresponding author
- Nieder, Andreas
- Venue
- Science Advances
- Published
- 2019-05-03
- Volume
- 5
- Issue
- 5
- Pages
- eaav7903-eaav7903
- ISSN
- 2375-2548 (ISSN-L)
- PMC
- PMC6506249
- OpenAlex
- W2944629072
- Cited by
- 110
- References
- 41
- Citations by year
- 2026: 4 · 2025: 11 · 2024: 18 · 2023: 19 · 2022: 19 · 2021: 20 · 2020: 15 · 2019: 4
- FWCI
- 11.34
- 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
- Deutsche Forschungsgemeinschaft (NI 618/10-1)
- Topics
- Social Sciences › Psychology › Developmental and Educational Psychology › Cognitive and developmental aspects of mathematical skills
- Life Sciences › Neuroscience › Cognitive Neuroscience › Visual perception and processing mechanisms
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Multisensory perception and integration
- Keywords
- number sense, numerosity, spontaneous emergence, deep neural network, visual object recognition, number discrimination, Weber-Fechner law, biologically inspired models, visual system
- MeSH terms
- Animals, Humans, Models, Biological, Neurons, Visual Perception, 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
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.