Gaussian RBF Centered Kernel Alignment (CKA) in the Large Bandwidth Limit
Not on the map. Why?
Abstract (arXiv)
We prove that Centered Kernel Alignment (CKA) based on a Gaussian RBF kernel converges to linear CKA in the large-bandwidth limit. We show that convergence onset is sensitive to the geometry of the feature representations, and that representation eccentricity bounds the range of bandwidths for which Gaussian CKA behaves nonlinearly.
Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 8 of its 29 references that are in the map. First posted on arXiv 17 Dec 2021.
Builds on
8 of 29 references.
References · 8
- 2014Adam: A Method for Stochastic Optimizationcited by 37,240
- 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,611
- 2019Similarity of Neural Network Representations Revisitedcited by 1,755
- 2014OpenML: networked science in machine learningcited by 526
- 2020Do Wide and Deep Networks Learn the Same Things? Uncovering How Neural Network Representations Vary with Width and Depthcited by 245
- 2019Emerging Cross-lingual Structure in Pretrained Language Modelscited by 181
- 2020Entangled Watermarks as a Defense against Model Extractioncited by 149
- 2019Universality and individuality in neural dynamics across large populations of recurrent networkscited by 61
Built on by
Citing papers per year: 2023: 1; 2024: 1; 2025: 2; 2026: 2.
6 papers.
- 2024Objective drives the consistency of representational similarity across datasetscited by 18
- 2025Lie Detector: Unified Backdoor Detection via Cross-Examination Frameworkcited by 6
- 2023Internal Cross-layer Gradients for Extending Homogeneity to Heterogeneity in Federated Learningcited by 5
- 2025Seeing Through the Brain: New Insights from Decoding Visual Stimuli with fMRIcited by 0
- 2026Data eccentricity, asymptotics of Gaussian RBF reproducing kernel Hilbert space, and kernel PCAcited by 0
- 2026DCRL: Decoupling and Coupling Reinforcement Learning via Policy-Reward Manifold Alignmentcited by 0
- 2026DCRL: Decoupling and Coupling Reinforcement Learning via Policy-Reward Manifold Alignment23 Sep 2026
- 2026Data eccentricity, asymptotics of Gaussian RBF reproducing kernel Hilbert space, and kernel PCA23 Jul 2026
- 2025Seeing Through the Brain: New Insights from Decoding Visual Stimuli with fMRI17 Oct 2025
- 2025Lie Detector: Unified Backdoor Detection via Cross-Examination Framework21 Mar 2025
- 2024Objective drives the consistency of representational similarity across datasets8 Nov 2024
- 2023Internal Cross-layer Gradients for Extending Homogeneity to Heterogeneity in Federated Learning22 Aug 2023