t-SNE, Forceful Colorings and Mean Field Limits
Not on the map. Why?
Abstract (arXiv)
t-SNE is one of the most commonly used force-based nonlinear dimensionality reduction methods. This paper has two contributions: the first is forceful colorings, an idea that is also applicable to other force-based methods (UMAP, ForceAtlas2,...). In every equilibrium, the attractive and repulsive forces acting on a particle cancel out: however, both the size and the direction of the attractive (or repulsive) forces acting on a particle are related to its properties: the force vector can serve as an additional feature. Secondly, we analyze the case of t-SNE acting on a single homogeneous cluster (modeled by affinities coming from the adjacency matrix of a random k-regular graph); we derive a mean-field model that leads to interesting questions in classical calculus of variations. The model predicts that, in the limit, the t-SNE embedding of a single perfectly homogeneous cluster is not a point but a thin annulus of diameter ~ k-1/4 n-1/4. This is supported by numerical results. The mean field ansatz extends to other force-based dimensionality reduction methods.
Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 7 of its 26 references that are in the map. First posted on arXiv 25 Feb 2021.
Builds on
7 of 26 references.
References · 7
- 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,841
- 2020Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMAP, and PaCMAP for Data Visualizationcited by 170
- 2016Visualizing Large-scale and High-dimensional Datacited by 79
- 2017Efficient Algorithms for t-distributed Stochastic Neighborhood Embeddingcited by 75
- 2017Clustering with t-SNE, provablycited by 51
- 2018An Analysis of the t-SNE Algorithm for Data Visualizationcited by 49
- 2019Heavy-tailed kernels reveal a finer cluster structure in t-SNE visualisationscited by 22
Built on by
Citing papers per year: 2021: 1; 2022: 2; 2023: 1; 2024: 5; 2025: 0; 2026: 1.
10 papers.
- 2023Manifold learning: what, how, and whycited by 66
- 2021Theoretical Foundations of t-SNE for Visualizing High-Dimensional Clustered Datacited by 55
- 2024Convergence analysis of t-SNE as a gradient flow for point cloud on a manifoldcited by 4
- 2024Large data limits and scaling laws for tSNEcited by 4
- 2024Assessing and improving reliability of neighbor embedding methods: a map-continuity perspectivecited by 4
- 2024Stochastic neighborhood embedding and the gradient flow of relative entropycited by 1
- 2024Attraction-Repulsion Swarming: A Generalized Framework of t-SNE via Force Normalization and Tunable Interactionscited by 1
- 2026On the continuum limit of t-SNE for data visualizationcited by 1
- 2022Structure from Voltagecited by 0
- 2022May the force be with youcited by 0
- 2026On the continuum limit of t-SNE for data visualization13 Apr 2026
- 2024Attraction-Repulsion Swarming: A Generalized Framework of t-SNE via Force Normalization and Tunable Interactions15 Nov 2024
- 2024Assessing and improving reliability of neighbor embedding methods: a map-continuity perspective22 Oct 2024
- 2024Large data limits and scaling laws for tSNE16 Oct 2024
- 2024Stochastic neighborhood embedding and the gradient flow of relative entropy25 Sep 2024
- 2024Convergence analysis of t-SNE as a gradient flow for point cloud on a manifold31 Jan 2024
- 2023Manifold learning: what, how, and why7 Nov 2023
- 2022May the force be with you13 Aug 2022
- 2022Structure from Voltage28 Feb 2022
- 2021Theoretical Foundations of t-SNE for Visualizing High-Dimensional Clustered Data16 May 2021