mlsanityData to 14 Oct 2026

t-SNE Exaggerates Clusters, Provably

arXiv 2510.07746 ↗First posted 9 Oct 2025cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

Central to the widespread use of t-distributed stochastic neighbor embedding (t-SNE) is the conviction that it produces visualizations whose structure roughly matches that of the input. To the contrary, we prove that (1) the strength of the input clustering, and (2) the extremity of outlier points, cannot be reliably inferred from the t-SNE output. We demonstrate the prevalence of these failure modes in practice as well.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 17 of its 40 references that are in the map. First posted on arXiv 9 Oct 2025.

Builds on

17 of 40 references.

References · 17

  1. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,372
  2. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,898
  3. 2016Visualizing Large-scale and High-dimensional Datacited by 79
  4. 2019TriMap: Large-scale Dimensionality Reduction Using Tripletscited by 60
  5. 2021Theoretical Foundations of t-SNE for Visualizing High-Dimensional Clustered Datacited by 55
  6. 2017Clustering with t-SNE, provablycited by 51
  7. 2018An Analysis of the t-SNE Algorithm for Data Visualizationcited by 49
  8. 2024Text Clustering with Large Language Model Embeddingscited by 27
  9. 2018Stochastic Neighbor Embedding under f-divergencescited by 9
  10. 2025Low-dimensional embeddings of high-dimensional datacited by 8
  11. 2017Stochastic Neighbor Embedding separates well-separated clusterscited by 7
  12. 2022Robust Inference of Manifold Density and Geometry by Doubly Stochastic Scalingcited by 7
5 more
  1. 2024Convergence analysis of t-SNE as a gradient flow for point cloud on a manifoldcited by 4
  2. 2021T-SNE Is Not Optimized to Reveal Clusters in Datacited by 3
  3. 2025Compressibility Barriers to Neighborhood-Preserving Data Visualizationscited by 3
  4. 2014Connection graph Laplacian methods can be made robust to noisecited by 2
  5. 2024Stochastic neighborhood embedding and the gradient flow of relative entropycited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026On the continuum limit of t-SNE for data visualizationcited by 1
  2. 2026On the Abundance of Critical Points of the t-SNE Energycited by 0
  3. 2026ibUMAP: Coherent and Scalable Field Evaluation for UMAP Optimizationcited by 0