The concept of "height" in number theory is a way of measuring the arithmetic complexity of points on geometric spaces called varieties. For certain well-behaved line bundles, called semipositive adelic line bundles, there is a notion of an "absolute minimum," which is the smallest height value that points on the variety can approach. A natural and long-standing question is whether this minimum is always actually achieved by some algebraic point, meaning a point whose coordinates satisfy polynomial equations with rational coefficients.
This paper answers that question with a surprising "no." The authors construct a concrete example of a semipositive adelic line bundle on a projective variety defined over the rational numbers where the absolute minimum exists as a numerical value, but no algebraic point on the variety actually achieves it. In other words, the infimum of the height function is approached arbitrarily closely but never quite reached by any algebraic point.
The key tool for building this counterexample is a carefully chosen family of dynamical systems on the projective line, meaning iterated maps from a geometric space to itself. The arithmetic properties of these maps, particularly how they behave at different prime numbers, allow the authors to engineer a height function with the desired gap between its theoretical minimum and what algebraic points can actually attain. This result refines the community's understanding of the relationship between height functions and algebraic points, showing that semipositivity alone is not strong enough to guarantee the minimum is achieved.