On cusp holonomies in strictly convex projective geometry
Balthazar Fléchelles
TL;DR
The paper provides a complete characterization of cusp holonomies in strictly convex and round convex projective geometry by establishing a criterion based on $P_1$-divergence and limit sets, and then extends the generalized cusp framework to include virtually solvable holonomies, yielding non-nilpotent examples. It develops a geometric and analytic toolkit—Hilbert geometry, duality, Vinberg hypersurfaces, and a smoothing lemma—to construct and deform cusp domains while preserving cusp structure. A key contribution is linking cusp holonomies to relatively Anosov representations, clarifying the periphery structure of these representations and enabling new constructions with virtually nilpotent or solvable holonomies. The work also demonstrates the existence of generalized cusps of non-maximal rank, including explicit solvable-holonomy examples, thereby highlighting the broader landscape of cusps in convex projective geometry and its connections to higher-rank dynamics.
Abstract
We give a complete characterization of the holonomies of strictly convex cusps and of round cusps in convex projective geometry. We build families of generalized cusps of non-maximal rank associated to each strictly convex or round cusp. We also extend Ballas-Cooper-Leitner's definition of generalized cusp to allow for virtually solvable fundamental group, and we produce the first such example with non-virtually nilpotent fundamental group. Along with a companion paper, this allows to build strictly convex cusps and generalized cusps whose fundamental group is any finitely generated virtually nilpotent group. This also has interesting consequences for the theory of relatively Anosov representations.
