Pith. sign in

REVIEW 4 major objections 5 minor 43 references

Mice to Machines: Neural Representations from Visual Cortex for Domain Generalization

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A brain-inspired normalization layer makes deep networks more robust to image-domain differences.

desk verdict NeuRN is effectively a local contrast normalization variant, and without that baseline the biological-DG link is untested; Table 1 is mixed and lacks statistics. read the letter →

arxiv 2505.06886 v1 pith:CW4B4Z7N submitted 2025-05-11 cs.CV cs.AIcs.LGcs.NE

Ahmed Qazi , Hamd Jalil , Asim Iqbal This is my paper · ORCID
classification cs.CVcs.AIcs.LGcs.NE
keywords domaingeneralizationmousevisualcortexneuralresponsenormalizationrepresentationalsimilaritybiologicallyinspirednetworkscalciumimagingcross-domaindigitclassificationcontrast
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to prove that deep networks become more robust to image-domain shifts when their internal feature representations are pushed toward the response patterns of the mouse visual cortex. It builds a comparison framework that matches feature maps from pretrained networks against calcium-imaging responses of excitatory and inhibitory neurons, and reports that excitatory neurons align more closely with network features. It then adds NeuRN, a normalization layer that divides each pixel's value by the standard deviation of its surrounding patch, and claims that NeuRN increases both biological similarity and cross-domain digit classification accuracy. If right, the paper offers a simple, parameter-free layer that any classifier can adopt.

What carries the argument

The load-bearing object is NeuRN, a normalization procedure that replaces each pixel's value by the inverted standard deviation of its local $k \times k$ patch, scaled by a global maximum; it is inserted into existing deep networks without learned parameters. The supporting comparison machinery is the RMSE between neural representations $N = T^T \cdot FS$ (trial traces transposed and multiplied by flattened stimuli) and reshaped feature maps, alongside kernel-density overlap measured by intersection-over-union. These comparisons are what the paper uses to argue that NeuRN increases biological alignment and that this alignment is the reason the models generalize.

What would settle it

A permutation test would settle it: shuffle the spatial locations within the neural representation matrices, or within the feature maps, and recompute RMSE to the other system. If shuffled alignments achieve RMSE as low as the intact ones, the reported biological similarity is an artifact of overall scale rather than spatial structure. A complementary test would compare randomly initialized or random-label-trained networks; if they match mouse visual cortex representations equally well, the alignment is not due to learned features.

Watch

Extended reading notes

Core claim

The paper's central claim is that the functional mapping of the mouse visual cortex closely tracks the feature representations of high-performing deep networks, especially for excitatory neurons, and that a deliberately inserted normalization layer, NeuRN, pushes artificial representations closer to biological ones while improving classification accuracy under domain shift. NeuRN computes, for each pixel, the standard deviation of its surrounding $k \times k$ patch and divides by a global maximum standard deviation to produce a contrast-normalized feature map. The authors evaluate this with two measurements: the RMSE between flattened neural and feature representations, and the overlap of kernel-density estimates of activation distributions. On four digit datasets, models with NeuRN often beat their base versions when trained on one style of digits and tested on another.

Load-bearing premise

The biological-alignment evidence depends on the assumption that a neuron's response traces, reshaped into the same grid as a network's feature map, have a meaningful spatial correspondence with those feature maps; if the time and pixel axes do not correspond to the spatial layout of features, the RMSE numbers do not establish biological alignment.

Editorial extensions

If this is right

  • Models with NeuRN should hold up better when test images differ in background, font, or lighting from training digits; the paper reports gains across twelve source-to-target transfer paths on four digit datasets.
  • NeuRN can be inserted into any CNN or Vision Transformer before standard training, since it requires no change to loss function or architecture.
  • The representational comparison pipeline can be applied to other neuron types, brain regions, or species to search for stronger candidate inductive biases.
  • If biological alignment is the mechanism, networks whose feature statistics match cortical responses more closely should continue to gain robustness as the alignment improves.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct extension would be to test NeuRN on harder natural distribution shifts, such as corrupted versions of standard benchmarks, to see whether local-contrast normalization helps beyond digit fonts and backgrounds.
  • The success of NeuRN suggests a testable hypothesis for neuroscience: local contrast normalization may be a major reason cortical representations stay stable across illumination and background changes; this could be checked with recordings under controlled contrast manipulations.
  • Because the paper compares one temporally integrated neural response to static feature maps, a stronger version of its claim would compare layer responses time-locked to stimulus onset; the reported RMSE alignment does not yet control for latency.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The manuscript proposes a framework for comparing neural representations from the mouse visual cortex (Allen Brain Observatory) with feature representations from deep networks, using RMSE and KDE-based measures. Based on an observed alignment, the authors introduce Neural Response Normalization (NeuRN), a biologically motivated local divisive normalization layer, and integrate it into several DNNs. They report that NeuRN increases biological alignment and improves domain generalization across digit datasets (MNIST, SVHN, USPS, MNIST-M), concluding that NeuRN captures domain-agnostic features. The paper also presents population-level and single-neuron-level comparisons across brain regions and genotypes.

Significance. If the claims were substantiated, the work would offer a simple, biologically inspired layer that improves robustness and strengthens the link between rodent visual cortex and DNN representations. The paper spans multiple architectures and brain regions, and it uses a public neuroscience dataset, which is a strength. However, the current evidence is not sufficient: the domain generalization table is mixed and lacks statistics, no generic normalization baselines are included, and the representational similarity metric relies on an unvalidated spatial correspondence. The absence of code and of the key hyperparameter (patch size k) further limits reproducibility.

major comments (4)
  1. [Table 1] The claim that NeuRN yields 'significant improvement' in domain generalization is not supported by the full table. For example, DenseNet121 drops from 74.3 to 26.4 on M→U, ResNet50v2 drops from 78.7 to 31.4 on M→U, and VGG19 drops from 66.2 to 48.4 on U→M, while improvements on other entries are large. No error bars or significance tests are reported for any entry, so the highlighted improvements cannot be distinguished from noise. The text selects four transfers (M→S, M→MM, U→S, MM→S) that improve, but this selection is post hoc. Please report mean and standard deviation over multiple seeds with paired significance tests and discuss the inconsistent pattern.
  2. [Methods, Neural Response Normalization (NeuRN)] NeuRN computes 1/(c·σ_pk) over k×k patches, which is a local divisive normalization in the same family as Local Contrast Normalization (Jarrett et al. 2009) and Local Response Normalization (Krizhevsky et al. 2012), both cited but never run as controls. Without a control experiment using LCN or LRN at matched capacity, any improvement in Table 1 could be attributed to generic normalization rather than to the biological inspiration. The biological-alignment results in Figures 2 and 3 need the same control before the claim that NeuRN specifically improves alignment can be accepted.
  3. [Methods, High-level representational analysis and Single neuron-level analysis] The RMSE comparison reshapes a neural representation N = T^T·FS, whose axes are time and pixels, and a DNN feature map, whose axes are spatial positions, into equal-length 1D vectors and computes raw RMSE. This assumes a meaningful spatial correspondence between the two modalities without any justification, and it is not the standard RSA methodology, which compares similarity structures rather than raw vectors. The KDE/IoU analysis (Eqs. 2–5) compares only marginal activation distributions, discarding spatial layout. Since the reported RMSE differences in Figure 2 are small (on the order of 0.01) and no error bars are given, the biological-alignment evidence for NeuRN is not established.
  4. [Results, NeuRN-derived performance boost for domain generalization in DNNs] The manuscript states that the domain generalization results 'validate our previous findings' (Iqbal et al. 2024). Because NeuRN is adapted from that prior work and the biological-alignment improvement is measured with the same RMSE/KDE framework used in the design of NeuRN, the new evidence provided here may be partly circular. Please clarify what independent evidence is added beyond the prior work, and specify whether the evaluation metric was used as an objective during the design of NeuRN.
minor comments (5)
  1. [Methods, Neural Response Normalization (NeuRN)] The patch size k in the definition of NeuRN is never reported in the main text or the Appendix; without this value the experiments cannot be reproduced. The definition 'c = max(σ)' is ambiguous (max over which set of patches or channels) and no small constant is mentioned to avoid division by zero in flat regions.
  2. [References] The reference for UMAP is given as 'Hastie and McMahan 2018', but UMAP is by McInnes et al. (2018); please correct the citation.
  3. [Results, NeuRN improves feature and neural representational alignment in DNNs] The text states that differences are 'insignificant' without reporting a statistical test; please add test results or remove the term.
  4. [Table 1 caption] The caption uses 'bold' for improvement and 'underlined' for close-to-benchmark scores, but the criteria for 'close-to-benchmark' are not defined, and the table would be clearer if it reported standard deviations.
  5. [Abstract and Introduction] The abstract and introduction claim 'significant improvements' from NeuRN while Table 1 contains many degradations; please align the wording with the evidence.

Circularity Check

1 steps flagged · score 4.0 of 10

NeuRN's biological premise is imported from and validated by the same authors' unreviewed preprint, but the core DG and alignment measurements are independent empirical tests.

  1. self citation load bearing [Introduction; Results & Discussion, 'Neural Response Normalization (NeuRN)' and 'NeuRN-derived performance boost for domain generalization in DNNs']
    "Inspired from the Winner-Takes-All (WTA) mechanism observed in neuronal circuits (Iqbal et al. 2024), we applied this approach to enable NeuRN—Neural Response Normalization—a biologically inspired technique designed to enhance both the domain generalization capabilities and biological alignment of DNNs. ... This improvement is particularly evident in transitions from MNIST-M to SVHN, MNIST to SVHN, MNIST to MNIST-M, and USPS to SVHN, highlighting NeuRN’s strong domain bridging capability and validates our previous findings (Iqbal et al. 2024)."

    The paper's sole cited basis for the WTA mechanism and for NeuRN itself is Iqbal et al. 2024, an unreviewed bioRxiv preprint sharing the corresponding author. The current paper then uses its Table 1 results to 'validate our previous findings' from that same preprint. The method's biological premise is therefore imported from, and confirmed by, the same self-citation loop rather than by an independent derivation or an external benchmark for the WTA claim. The DG numbers are not mathematically forced by the citation, so this is a justification-level circularity, not a computational one. It also means the 'biologically inspired' explanation is not distinguished from generic local divisive normalization, since no LCN/LRN control is run.

full rationale

The core quantitative comparisons are not circular in a computational sense. The neural representation N=T^T·FS, the RMSE comparison (Eq. 1), and the KDE/IoU analysis are defined explicitly and evaluated against external Allen Brain Observatory data; no parameter of NeuRN is fitted to minimize RMSE or to force the reported DG accuracies. The domain-generalization experiments use held-out target domains (MNIST, SVHN, USPS, MNIST-M) and are therefore independent evidence, even if Table 1 is mixed and lacks error bars and significance tests. The main circular element is the self-referential provenance of NeuRN: the WTA inspiration and the validation of 'previous findings' both point to the same unreviewed preprint by the same team. That self-citation is load-bearing for the biological framing, though not for the raw empirical outcomes. The absence of a generic-normalization control (e.g., LCN or LRN) is a legitimate correctness/control concern rather than a circularity, so it does not raise the score above 4. The paper itself also acknowledges limitations—excitatory/inhibitory differences are 'insignificant' and NeuRN struggles in VISpm—which further supports a modest, non-inflated circularity score.

Assumptions & free parameters 4 free parameters · 5 assumptions · 1 invented entities

The central claims rest on several hand-picked parameters and domain assumptions. The most important free parameter is NeuRN's patch size k, never reported, which directly controls the normalization. The UMAP/K-means selection of neurons and features is a data-dependent choice that can bias the alignment comparison. The key axiom is that raw RMSE between reshaped neural and DNN tensors is a valid measure of representational similarity, which the paper does not justify.

free parameters (4)
  • NeuRN patch size k = not reported
    Control the scale of local contrast normalization in the NeuRN layer (Methods, 'Neural Response Normalization'). Essential for reproducing results, never stated.
  • UMAP n_neighbors = 15
    Hyperparameter for neuron and feature selection via UMAP (Methods, 'Neural Representations dataset').
  • UMAP min_dist = 0.1
    Hyperparameter for UMAP selection (Methods).
  • K-means number of clusters = 10
    Number of clusters used to select 'central' neural and feature representations (Methods). This selection determines which data enter the comparison.
assumptions (5)
  • domain assumption DNN feature maps and MVC neural representations can be validly compared via raw RMSE after reshaping to equal-length 1D vectors.
    Methods, 'High-level representational analysis', Eq. (1). The paper assumes a meaningful correspondence between the time×pixel axes of N=T^T·FS and the spatial axes of a conv filter channel, without justification.
  • domain assumption Fine-tuning on the CIFAR-27 subset transfers to the Allen natural-scene stimuli domain.
    Methods, 'CIFAR-27' section. The paper asserts the subset is close to natural scenes but provides no quantitative similarity measure.
  • ad hoc to paper NeuRN's local standard deviation normalization produces domain-agnostic features.
    Methods, 'Neural Response Normalization'. No derivation or proof of domain invariance is provided.
  • domain assumption Digit transfer tasks (M, S, U, MM) form a valid domain generalization benchmark for ImageNet-pretrained models.
    Evaluation datasets section. Near-chance baselines for several models (e.g., EfficientNetB0 M→U 8.6%) suggest the setup may not be well calibrated.
  • domain assumption UMAP+K-means selection of 'central' neurons and features yields an unbiased sample.
    Methods, 'Neural Representations dataset'. Selecting cluster centers could bias the comparison toward typical examples and inflate alignment.
invented entities (1)
  • NeuRN (Neural Response Normalization layer)
    purpose: Biologically inspired normalization layer intended to improve domain generalization and increase similarity between DNN features and mouse V1 responses.
    No external validation beyond the paper's own experiments and the authors' prior preprint (Iqbal et al. 2024). No public code or independent replication. The layer is structurally similar to Local Contrast Normalization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mice to Machines: Neural Representations from Visual Cortex for Domain Generalization." pith.science (2026). https://pith.science/paper/CW4B4Z7N

@misc{pith2026250506886,
  author       = {Pith},
  title        = {Pith review of: Mice to Machines: Neural Representations from Visual Cortex for Domain Generalization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CW4B4Z7N}},
  note         = {Machine review of arXiv:2505.06886}
}
read the original abstract

The mouse is one of the most studied animal models in the field of systems neuroscience. Understanding the generalized patterns and decoding the neural representations that are evoked by the diverse range of natural scene stimuli in the mouse visual cortex is one of the key quests in computational vision. In recent years, significant parallels have been drawn between the primate visual cortex and hierarchical deep neural networks. However, their generalized efficacy in understanding mouse vision has been limited. In this study, we investigate the functional alignment between the mouse visual cortex and deep learning models for object classification tasks. We first introduce a generalized representational learning strategy that uncovers a striking resemblance between the functional mapping of the mouse visual cortex and high-performing deep learning models on both top-down (population-level) and bottom-up (single cell-level) scenarios. Next, this representational similarity across the two systems is further enhanced by the addition of Neural Response Normalization (NeuRN) layer, inspired by the activation profile of excitatory and inhibitory neurons in the visual cortex. To test the performance effect of NeuRN on real-world tasks, we integrate it into deep learning models and observe significant improvements in their robustness against data shifts in domain generalization tasks. Our work proposes a novel framework for comparing the functional architecture of the mouse visual cortex with deep learning models. Our findings carry broad implications for the development of advanced AI models that draw inspiration from the mouse visual cortex, suggesting that these models serve as valuable tools for studying the neural representations of the mouse visual cortex and, as a result, enhancing their performance on real-world tasks.

Figures

Figures reproduced from arXiv: 2505.06886 by the authors.

Figure 1
Figure 1. Block diagram architecture of comparison framework: image stimuli are presented to mice and DNN models (step 1) [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. A) RMSE score comparison of feature representations from NeuRN and non-NeuRN DNN models with MVC neural [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Distributional comparison of NeuRN and non-NeuRN feature representations with neural representations is shown. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 8 canonical work pages

  1. [1]

    Allen Institute for Brain Science . 2016. Brain Observatory. https://observatory.brain-map.org/visualcoding/stimulus/natural_scenes

    work page 2016

  2. [2]

    Ben-David, S.; Blitzer, J.; Crammer, K.; Kulesza, A.; Pereira, F.; and Vaughan, J. W. 2010. A theory of learning from different domains. In Proceedings of the 27th International Conference on Machine Learning (ICML-10), 101--108. ACM

    2010

  3. [3]

    A.; Dogan, U.; Lee, G.; and Scott, C

    Blanchard, G.; Deshmukh, A. A.; Dogan, U.; Lee, G.; and Scott, C. 2021. Domain generalization by marginal transfer learning. Journal of Machine Learning Research, 22(145): 1--41

    2021

  4. [4]

    F.; et al

    Burg, M. F.; et al. 2021. Learning divisive normalization in primary visual cortex. PLOS Computational Biology, 17(6): e1009028

    2021

  5. [5]

    A.; Denfield, G

    Cadena, S. A.; Denfield, G. H.; Walker, E. Y.; Gatys, L. A.; Tolias, A. S.; Bethge, M.; and Ecker, A. S. 2019 a . Deep convolutional models improve predictions of macaque V1 responses to natural images. PLoS computational biology, 15(4): e1006897

    2019

  6. [6]

    A.; Sinz, F

    Cadena, S. A.; Sinz, F. H.; Muhammad, T.; Froudarakis, E.; Cobos, E.; Walker, E. Y.; Reimer, J.; Bethge, M.; Tolias, A. S.; and Ecker, A. S. 2019 b . How well do deep neural networks trained on object recognition characterize the mouse visual system

    2019

  7. [7]

    F.; Hong, H.; Yamins, D

    Cadieu, C. F.; Hong, H.; Yamins, D. L.; Pinto, N.; Ardila, D.; Solomon, E. A.; Majaj, N. J.; and DiCarlo, J. J. 2014. Deep neural networks rival the representation of primate IT cortex for core visual object recognition. PLoS Computational Biology, 10(12): e1003963

    2014

  8. [8]

    Chollet, F. 2017. Xception: Deep Learning with Depthwise Separable Convolutions. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1800--1807

    2017

Show all 43 references
  1. [9]

    Conwell, C.; et al. 2021. Neural Regression, Representational Similarity, Model Zoology & Neural Taskonomy at Scale in Rodent Visual Cortex. In NeurIPS Proceedings

    2021

  2. [10]

    Das, A.; and Ray, S. 2021. Effect of Cross-Orientation Normalization on Different Neural Measures in Macaque Primary Visual Cortex. Cerebral Cortex Communications, 2(1): tgab009

    2021

  3. [11]

    Ganin, Y.; Ustinova, E.; Ajakan, H.; Germain, P.; Larochelle, H.; Laviolette, F.; Marchand, M.; and Lempitsky, V. 2015. Domain-Adversarial Training of Neural Networks

    2015

  4. [12]

    Gao, R.; Zou, H.; and Duan, Z. 2023. CVSNet: A Computer Implementation for Central Visual System of The Brain. arXiv:2305.19492

    2023 arXiv

  5. [13]

    Hastie, T.; and McMahan, B. J. 2018. UMAP: Uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426

    2018 arXiv

  6. [14]

    He, K.; Zhang, X.; Ren, S.; and Sun, J. 2016. Deep Residual Learning for Image Recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770--778

    2016

  7. [15]

    Hull, J. J. 1994. A database for handwritten text recognition research. IEEE Transactions on pattern analysis and machine intelligence, 16(5): 550--554

    1994

  8. [16]

    Iqbal, A.; Dong, P.; Kim, C.; and Jang, H. 2019. Decoding Neural Responses in Mouse Visual Cortex through a Deep Neural Network. In 2019 International Joint Conference on Neural Networks (IJCNN). IEEE

    2019

  9. [17]

    J.; Fishell, G.; and Honnuraiah, S

    Iqbal, A.; Mahmood, H.; Stuart, G. J.; Fishell, G.; and Honnuraiah, S. 2024. Biologically Realistic Computational Primitives of Neocortex Implemented on Neuromorphic Hardware Improve Vision Transformer Performance. bioRxiv. This article is a preprint and has not been certified...

    2024

  10. [18]

    Jarrett, K.; Kavukcuoglu, K.; Ranzato, M.; and LeCun, Y. 2009. What is the best multi-stage architecture for object recognition? 2009 IEEE 12th International Conference on Computer Vision, 2146--2153

    2009

  11. [19]

    Jozwik, K.; et al. 2019. To find better neural network models of human vision, find better neural network models of primate vision. DiCarlo Lab

    2019

  12. [20]

    Kar, K.; et al. 2019. Evidence that recurrent circuits are critical to the ventral stream’s execution of core object recognition behavior. Nature Neuroscience, 22: 974--983

    2019

  13. [21]

    Khaligh-Razavi, S.-M.; and Kriegeskorte, N. 2014. Deep supervised, but not unsupervised, models may explain IT cortical representation. PLoS Computational Biology, 10(11): e1003915

    2014

  14. [22]

    Krizhevsky, A.; Sutskever, I.; and Hinton, G. 2012. ImageNet Classification with Deep Convolutional Neural Networks. In Neural Information Processing Systems, volume 25

    2012

  15. [23]

    L.; and DiCarlo, J

    Kubilius, J.; Schrimpf, M.; Nayebi, A.; Bear, D.; Yamins, D. L.; and DiCarlo, J. J. 2018. CORnet: Modeling the Neural Mechanisms of Core Object Recognition. bioRxiv

    2018

  16. [24]

    Kubilius, J.; et al. 2019. Brain-Like Object Recognition with High-Performing Shallow Recurrent ANNs. In NeurIPS Proceedings

    2019

  17. [25]

    LeCun, Y.; Cortes, C.; and Burges, C. 2010. MNIST Handwritten Digit Database

    2010

  18. [26]

    Li, Z.; et al. 2019. Learning from brains how to regularize machines. In Advances in Neural Information Processing Systems 32 (NeurIPS 2019)

    2019

  19. [27]

    Markram, H.; Toledo-Rodriguez, M.; Wang, Y.; Gupta, A.; Silberberg, G.; and Wu, C. 2004. Interneurons of the neocortical inhibitory system. Nature Reviews Neuroscience, 5(10): 793--807

    2004

  20. [28]

    G.; Raeder, T.; Alaiz-Rodr ' guez, R.; Chawla, N

    Moreno-Torres, J. G.; Raeder, T.; Alaiz-Rodr ' guez, R.; Chawla, N. V.; and Herrera, F. 2012. A unifying view on dataset shift in classification. Pattern Recognition, 45(1): 327--336

    2012

  21. [29]

    Netzer, Y.; Wang, T.; Coates, A.; Bissacco, A.; Wu, B.; and Ng, A. Y. 2011. Reading Digits in Natural Images with Unsupervised Feature Learning. In NIPS Workshop on Deep Learning and Unsupervised Feature Learning

    2011

  22. [30]

    Recht, B.; Roelofs, R.; Schmidt, L.; and Shankar, V. 2019. Do imagenet classifiers generalize to imagenet? In International Conference on Machine Learning, 5401--5410

    2019

  23. [31]

    Sawada, T.; and Petrov, A. 2016. Emulating and predicting physiological results of neurons in the primary visual cortex (V1) based on the divisive normalization model. Journal of Vision, 16(12): 958

    2016

  24. [32]

    B.; Kar, K.; Bashivan, P.; Prescott-Roy, J.; Schmidt, K.; Yamins, D

    Schrimpf, M.; Kubilius, J.; Hong, H.; Majaj, N.; Rajalingham, R.; Issa, E. B.; Kar, K.; Bashivan, P.; Prescott-Roy, J.; Schmidt, K.; Yamins, D. L.; and DiCarlo, J. J. 2018. Brain-Score: Which Artificial Neural Network for Object Recognition is most Brain-Like? bioRxiv

    2018

  25. [33]

    Shi, J.; Shea-Brown, E.; and Buice, M. 2019. Comparison against task driven artificial neural networks reveals functional properties in mouse visual cortex. Advances in Neural Information Processing Systems, 32

    2019

  26. [34]

    Shi, J.; et al. 2022. MouseNet: A biologically constrained convolutional neural network model for the mouse visual cortex. PLOS Computational Biology

    2022

  27. [35]

    Simonyan, K.; and Zisserman, A. 2015. Very Deep Convolutional Networks for Large-Scale Image Recognition. In 3rd International Conference on Learning Representations (ICLR 2015), 1--14

    2015

  28. [36]

    Taori, R.; Dave, A.; Shankar, V.; Carlini, N.; Recht, B.; and Schmidt, L. 2020. Measuring robustness to natural distribution shifts in image classification. In Advances in Neural Information Processing Systems, volume 33, 10581--10591

    2020

  29. [37]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. In Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS'17), 6000--6010. Curran Associates Inc

    2017

  30. [38]

    Veerabadran, V.; Raina, R.; and Sa, V. D. ???? Bio-inspired learnable divisive normalization for ANNs. No additional publication details available

  31. [39]

    H.; Kunz, E

    Williams, A. H.; Kunz, E. M.; Kornblith, S.; and Linderman, S. W. 2021. Generalized Shape Metrics on Neural Representations

    2021

  32. [40]

    L.; Hong, H.; Cadieu, C

    Yamins, D. L.; Hong, H.; Cadieu, C. F.; Solomon, E. A.; Seibert, D.; and DiCarlo, J. J. 2014. Performance-optimized hierarchical models predict neural responses in higher visual cortex. Proceedings of the National Academy of Sciences of the United States of America, 111(23): 8...

    2014

  33. [41]

    A.; Maunsell, J

    Zayyad, Z. A.; Maunsell, J. H. R.; and MacLean, J. N. 2023. Normalization in mouse primary visual cortex. bioRxiv. The complete version history of this preprint is available at bioRxiv

    2023

  34. [42]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  35. [43]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.