Estimating Neural Representation Alignment from Sparsely Sampled Inputs and Features
Open accessAbstract
In both artificial and biological systems, the centered kernel alignment (CKA) has become a widely used tool for quantifying neural representation similarity. While current CKA estimators typically correct for the effects of finite stimuli sampling, the effects of sampling a subset of neurons are overlooked, introducing notable bias in standard experimental scenarios. Here, we provide a theoretical analysis showing how this bias is affected by the representation geometry. We then introduce a novel estimator that corrects the bias for both input and feature sampling. We use our method for evaluating both brain-to-brain and model-to-brain alignments and show that it delivers reliable comparisons even with very sparsely sampled neurons. We perform within-animal and across-animal comparisons on electrophysiological data from visual cortical areas V1, V4, and IT, and use these as benchmarks to evaluate model-to-brain alignment. We also apply our method to reveal how object representations become progressively disentangled across layers in both biological and artificial systems. These findings underscore the importance of correcting feature-sampling biases in CKA and demonstrate that our bias-corrected estimator provides a more faithful measure of representation alignment. The improved estimates increase our understanding of how neural activity is structured across both biological and artificial systems.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
In finite-sample, high-dimensional regimes, raw similarity scores can be systematically biased.
“In finite-sample, high-dimensional regimes, raw similarity scores can be systematically biased.”
From Revisiting the Platonic Representation Hypothesis: An Aristotelian View (2026) · cites this paper as Chun et al., 2025
The dep-cols CKA of Chun et al. (2025) is designed to correct column dependence, which is not a confound in the experimental setup (where columns are independent by construction), and as a result it attenuates the true signal under H1.
“In contrast, dep-cols CKA is designed to correct column dependence, which is not a confound in our experimental setup (where columns are independent by construction), and as a result, it attenuates the true signal under H 1 .”
From Revisiting the Platonic Representation Hypothesis: An Aristotelian View (2026) · cites this paper as Chun et al. (2025)
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Still being checked.
Research works printed but not linked: 24
- Cadena, S. A., Willeke, K. F., Restivo, K., Denfield, G., Sinz, F. H., Bethge, M., . . . Ecker, A. S. (2024). Diverse task- driven modeling of macaque v4 reveals functional specializa- tion towards semantic tasks. PLOS Computational Biology , 20(5), e1012056.
- Cloos, N., Siegel, M., Brincat, S. L., Miller, E. K., & Cueva, C. J. (2024). Differentiable optimization of similarity scores Proceedings of Cognitive Computational Neuroscience 2025 between models and brains. In Iclr 2024 workshop on repre- sentational alignment.
- Davari, M., Horoi, S., Natik, A., Lajoie, G., Wolf, G., & Belilovsky, E. (2022). Reliability of cka as a similarity measure in deep learning. arXiv preprint arXiv:2210.16156. · arXiv
- Gretton, A., Bousquet, O., Smola, A., & Schölkopf, B. (2005). Measuring statistical dependence with hilbert-schmidt norms. In International conference on algorithmic learning theory (pp. 63-77).
- Han, Y., Poggio, T. A., & Cheung, B. (2023). System identifi- cation of neural systems: If we got it right, would we know? In International conference on machine learning (pp. 12430- 12444).
- He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the ieee conference on computer vision and pattern recognition (pp. 770-778).
- Hebart, M. N., Dickter, A. H., Kidder, A., Kwok, W. Y., Corriveau, A., Van Wicklin, C., & Baker, C. I. (2019). Things: A database of 1,854 object concepts and more than 26,000 naturalistic object images. PloS one, 14(10), e0223792.
- Hoeffding, W. (1948). A class of statistics with asymptotically normal distribution. The Annals of Mathematical Statistics, 19(3), 293-325.
- Hubel, D. H., & Wiesel, T. N. (1962). Receptive fields, binocu- lar interaction and functional architecture in the cat's visual cortex. The Journal of physiology , 160(1), 106.
- Hung, C. P., Kreiman, G., Poggio, T., & DiCarlo, J. J. (2005). Fast readout of object identity from macaque inferior tempo- ral cortex. Science, 310(5749), 863-866.
- Kornblith, S., Norouzi, M., Lee, H., & Hinton, G. (2019). Simi- larity of neural network representations revisited. In Interna- tional conference on machine learning (pp. 3519-3529).
- Kriegeskorte, N., & Kievit, R. A. (2013). Representational geometry: integrating cognition, computation, and the brain. Trends in cognitive sciences, 17 (8), 401-412.
- Kubilius, J., Schrimpf, M., Nayebi, A., Bear, D., Yamins, D. L., & DiCarlo, J. J. (2018). Cornet: Modeling the neural mecha- nisms of core object recognition. BioRxiv , 408385.
- Majaj, N. J., Hong, H., Solomon, E. A., & DiCarlo, J. J. (2015). Simple learned weighted sums of inferior temporal neuronal firing rates accurately predict human core object recogni- tion performance. Journal of Neuroscience, 35(39), 13402- 13418.
- Murphy, A., Zylberberg, J., & Fyshe, A. (2024). Correcting biased centered kernel alignment measures in biological and artificial neural networks. arXiv preprint arXiv:2405.01012. · arXiv
- Nguyen, T., Raghu, M., & Kornblith, S. (2021). Do wide and deep networks learn the same things? uncovering how neural network representations vary with width and depth. In International conference on learning representa- tions. Retrieved from https://openreview.net/forum ?id=KJNcAkY8tY4
- Raghu, M., Unterthiner, T., Kornblith, S., Zhang, C., & Dosovit- skiy, A. (2021). Do vision transformers see like convolutional neural networks? Advances in neural information processing systems, 34, 12116-12128.
- Schrimpf, M., Kubilius, J., Hong, H., Majaj, N. J., Rajalingham, R., Issa, E. B., . . . others (2018). Brain-score: Which artificial neural network for object recognition is most brain- like? BioRxiv , 407007.
- Song, L., Smola, A., Gretton, A., Bedo, J., & Borgwardt, K. (2012). Feature selection via dependence maximization. Journal of Machine Learning Research, 13(5).
- Stringer, C., Pachitariu, M., Steinmetz, N., Carandini, M., & Har- ris, K. D. (2019). High-dimensional geometry of population responses in visual cortex. Nature, 571(7765), 361-365.
- Sucholutsky, I., Muttenthaler, L., Weller, A., Peng, A., Bobu, A., Kim, B., . . . others (2023). Getting aligned on representa- tional alignment. arXiv preprint arXiv:2310.13018. · arXiv
- Williams, A. H., Kunz, E., Kornblith, S., & Linderman, S. (2021). Generalized shape metrics on neural representations. Ad- vances in Neural Information Processing Systems, 34, 4738- 4750.
- Yamins, D. L., & DiCarlo, J. J. (2016). Using goal-driven deep learning models to understand sensory cortex. Nature neuroscience, 19(3), 356-365.
- Yamins, D. L., Hong, H., Cadieu, C. F., Solomon, E. A., Seibert, D., & DiCarlo, J. J. (2014). Performance-optimized hierarchi- cal models predict neural responses in higher visual cortex. Proceedings of the national academy of sciences, 111(23), 8619-8624. Proceedings of Cognitive Computational Neuroscience 2025
Cited by
Sorted by citations
Revisiting the Platonic Representation Hypothesis: An Aristotelian View
Gröger, F., Wen, S., & Brbić, M. (2026) · arXiv · doi:10.48550/arxiv.2602.14486
“Figure 12 shows the difference between our calibrated CKA and two existing estimators: the debiased CKA of Murphy et al. (2024) and the dep-cols CKA of Chun et al. (2025).”
- Authors
- Chun, C., Canatar, A., Chung, S., & Lee, D.
- Affiliations
- Flatiron Institute (US)
- Flatiron Institute
- Harvard University (US)
- Cornell University (US)
- Venue
- Proceedings of Cognitive Computational Neuroscience 2025
- Publisher
- Cognitive Computational Neuroscience
- Published
- 2025
- OpenAlex
- W4416507587
- Cited by
- 0
- References
- 0
- FWCI
- 0.00
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference on Cognitive Computational Neuroscience (CCN) (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Samsung Advanced Institute of Technology
- Flatiron Health
- Ministry of Science and ICT, South Korea
- Samsung
- Institute for Information and Communications Technology Promotion
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › Visual perception and processing mechanisms
- Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Keywords
- centered kernel alignment, sampling bias correction, feature sampling, estimation bias, visual cortex, hierarchical visual processing, electrophysiology
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
