The paper studies when algebraic curves (one-dimensional geometric objects defined by polynomial equations) can be embedded inside special spaces called toric varieties. Toric varieties are a well-studied and relatively structured class of geometric spaces that include familiar examples like ordinary projective spaces and weighted projective spaces. The central question is: given a curve that fits inside some projective space of dimension d, must it also fit inside a simpler or lower-dimensional toric variety?
The main result is a negative answer to this question. For every integer d at least 2, the authors construct a specific curve that can be embedded in d-dimensional projective space but cannot be embedded in any projective normal toric variety of dimension smaller than d. This means these curves genuinely require the full dimension of the ambient projective space and cannot be compressed into a more economical toric setting. A particularly striking special case is that some curves live in d-dimensional projective space but cannot be embedded in any weighted projective space of dimension less than d, even though weighted projective spaces are quite flexible generalizations of ordinary projective space.
The obstructions the authors identify are rooted in the combinatorial and algebraic structure that toric varieties must satisfy. Toric varieties are constrained by the fact that they carry a large group of symmetries described by combinatorial data, and curves with sufficiently complicated singularities or numerical invariants can fail to be compatible with those constraints. By carefully engineering curves whose local structure at singular points conflicts with any low-dimensional toric ambient space, the authors show that dimension is a genuine and unavoidable resource when trying to embed such curves in this structured class of spaces.