A variational framework for modal estimation
Tâm Le Minh, Julyan Arbel, Florence Forbes, Hien Duy Nguyen · 2026
Read A variational framework for modal estimation by Tâm Le Minh, Julyan Arbel, Florence Forbes, Hien Duy Nguyen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-17956.txt.
Introduction
Tâm LeMinh and team introduce GERVE, Gibbs-measure Entropy-Regularized Variational Estimation, a sample-based method for multivariate mode estimation. Mode gauging arises in inverse problems, multi-mode sampling and density-based grouping. Yet it grows hard in mid to high dimensions, above all when the density cannot be judged.
GERVE approximates Gibbs distributions from samples without judging the density. It uses Gaussian-mixture variational cooling and natural tuning. That produces a mixture in high-density regions whose component shares also cluster data. The team proves theory in two regimes. As Gibbs temperature falls to zero, the best mixture gathers around global modes of the population density. At fixed positive heat, they prove existence and long-run bell shape of empirical winners and propose a bootstrap method for uncertainty scoring. Tests and a real-data test show GERVE recovers modes well and produces useful clusters.
Catalog record
| CREATOR | Tâm Le Minh, Julyan Arbel, Florence Forbes, Hien Duy Nguyen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 219 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-17956