Fixed-degree kernels over a number field

Determine which rational polygons P have a nonempty K-coefficient arithmetic kernel Ker_d(P,K) for a fixed number field K and degree bound d; determine whether the number of vertices or reflex vertices can bound a sufficient degree, and compare this kernel with the set of rational guards that see every real point of P along real-coefficient carriers of degree at most d.

Background

The arithmetic kernel consists of rational guards that can see every target using carriers of bounded degree over a specified coefficient field K, with no coefficient-height bound. At degree one, the K-coefficient kernel coincides with the classical kernel intersected with the rational points, whereas quadratic carriers can enlarge it, as demonstrated by the L-shaped gallery example.

The paper also shows that rational carriers may fail to provide a whole-gallery guard at a fixed degree, even in convex polygons, because algebraic targets can have relation degree exceeding the permitted carrier degree. The problem seeks structural criteria for nonemptiness and quantitative bounds in terms of polygonal complexity, together with a comparison to the analogous real-coefficient visibility notion.

References

Problem 10.2 (Fixed-degree kernels over a number field). Which rational polygons have nonempty Kerd(P, K)? For fixed K, can the number of vertices or reflex vertices bound a sufficient degree? How does this kernel compare with the set of rational guards that see every real point of P along real-coefficient carriers of degree at most d?

— Arithmetic Visibility in Polygons: Curved Sight in the Art Gallery  (2610.19506 - Barrett, 14 Oct 2026) in Problem 10.2, Section 10.2

Problem 10.5 (Singularities and admissible sight arcs). For the same arithmetic gallery, compare guard numbers for absolutely irreducible carriers under four rules: (i) the entire affine carrier is nonsingular; (ii) the sight arc avoids singularities, as here; (iii) the sight arc lifts continuously to the real normalization; (iv) connected full traces permit branch switching at singularities. How large can the differences be, and what budget increase replaces singular sight by regular sight?

— Arithmetic Visibility in Polygons: Curved Sight in the Art Gallery  (2610.19506 - Barrett, 14 Oct 2026) in Problem 10.5, Section 10.4

Whether the same guard sees every quadratic point of the upper arm on a rational conic remains a further question; see Problem 10.2.

— Arithmetic Visibility in Polygons: Curved Sight in the Art Gallery  (2610.19506 - Barrett, 14 Oct 2026) in Section 5, immediately after Example 5.5; revisited in Section 10.2