Leinster–Roff maximum-diversity support conjecture
Let be a Euclidean ball, equipped with the similarity kernel . A maximum-diversity probability measure on is a probability measure maximizing the diversity associated with . The conjecture asserts that, for every and , there exists a maximum-diversity measure whose support is a finite union of concentric spheres, namely for some and .
References
- Maximum Diversity and Support Transitions on Disks and Three-Dimensional Balls — arXiv — Adam Dajani
Source: arXiv
Progress summary
An unrefereed preprint claims the conjecture is false in dimensions two and three, while the higher-dimensional picture remains uncertain.
Leinster and Roff asked whether a maximum-diversity measure on a Euclidean ball is supported on finitely many concentric spheres. Their paper presents this as an open question.
Known results
- Leinster and Roff (2019; revised 2020): existence of maximum-diversity measures on compact metric spaces and characterization via positive weight measures.
- Leinster and Roff (2019; revised 2020): no support description was known for Euclidean balls or cubes; the finite-union-of-concentric-spheres description was conjectural.
October 2026 support-transition claim
Adam Dajani's unrefereed preprint claims that the maximizing measure is boundary-uniform below a dimension-dependent threshold, but above it is supported on a central ball and the outer sphere. This gives explicit measures and equations and claims to disprove the finitely-many-spheres conjecture in dimensions two and three; higher-dimensional behavior is described only computationally and conjecturally.
Current status (as of October 2026): The original support question is open in general; Dajani's claimed resolution in dimensions two and three is unverified, and higher-dimensional behavior remains conjectural.
Solutions 0
No solutions have been posted yet.