On the Abundance of Critical Points of the t-SNE Energy
Abstract
This paper considers the energy landscape of the t-SNE algorithm. While this algorithm has enjoyed broad adoption, the non-convexity of the associated energy has made it difficult to rigorously understand what the algorithm captures in many settings. In particular, a number of well-known numerical examples, several of which are reproduced in this article, suggest a complicated energy landscape with many local minimizers that do not respect the topology or clustering structure of the underlying data. This work seeks to provide first steps towards a rigorous explanation of these phenomena. Specifically, for a general family of energies, which include both the original t-SNE algorithm and recently identified large data limits, and for densities in feature space which obey a continuous symmetry, we construct infinite families of distinct critical points. These critical points are based upon identifying pairs of discrete symmetries, one in the original feature space and the other in the target embedding space, which are preserved under gradient dynamics. These critical configurations exhibit many characteristics, such as topology breaking and spurious clustering, which are often observed empirically. Finally, numerical and analytical examples are given throughout as a means of illustrating the approach.
Related
- One-bubble solution with maximum points escaping to infinity for the energy-critical heat equation in ℝ5
- A priori Dirichlet energy bounds for positive Trudinger--Moser critical points on simply connected planar domains
- A Second-Order Method for Locating Critical Points of Prescribed Index on Energy Landscapes
- On the Stability of Critical Points of the Hardy-Littlewood-Sobolev Inequality
- Response to the Letter to the Editor Entitled “A Critical Appraisal of End Points and Statistical Strategies in the DIDo Trial”
BibTeX
@misc{haridas2026abundance,
title = {{On the Abundance of Critical Points of the t-SNE Energy}},
author = {Nakul Haridas and Ryan Murray},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.04379},
eprint = {2609.04379},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.04379},
}
RIS
TY - UNPB
TI - On the Abundance of Critical Points of the t-SNE Energy
AU - Nakul Haridas
AU - Ryan Murray
PY - 2026
DA - 2026/09/03
PB - arXiv
DO - 10.48550/arxiv.2609.04379
UR - https://arxiv.org/abs/2609.04379
AB - This paper considers the energy landscape of the t-SNE algorithm. While this algorithm has enjoyed broad adoption, the non-convexity of the associated energy has made it difficult to rigorously understand what the algorithm captures in many settings. In particular, a number of well-known numerical examples, several of which are reproduced in this article, suggest a complicated energy landscape with many local minimizers that do not respect the topology or clustering structure of the underlying data. This work seeks to provide first steps towards a rigorous explanation of these phenomena. Specifically, for a general family of energies, which include both the original t-SNE algorithm and recently identified large data limits, and for densities in feature space which obey a continuous symmetry, we construct infinite families of distinct critical points. These critical points are based upon identifying pairs of discrete symmetries, one in the original feature space and the other in the target embedding space, which are preserved under gradient dynamics. These critical configurations exhibit many characteristics, such as topology breaking and spurious clustering, which are often observed empirically. Finally, numerical and analytical examples are given throughout as a means of illustrating the approach.
ER -