Hypersurfaces passing through the Galois orbit of a point

1. Problem Statement and Context

The paper addresses a fundamental question in arithmetic geometry regarding the existence of points in projective space over a field extension that avoid certain algebraic constraints defined over the base field.

• The Core Question: Let $K$ be a field and $L/K$ a separable extension of degree $r$. Let $n, d$ be positive integers. Does there exist a point $P \in \mathbb{P}^n(L)$ such that the Galois orbit of $P$ (under $\text{Gal}(L/K)$) is in "general position" with respect to the space of degree $d$ hypersurfaces defined over $K$?
• Formal Definition: Let $S_{n,d}(K)$ be the vector space of homogeneous polynomials of degree $d$ in $n+1$ variables over $K$, with dimension $m = \binom{n+d}{n}$. A point $P$ imposes linear constraints on this space. The authors seek a point $P$ such that the dimension of the subspace of forms vanishing at $P$ (and thus on its entire Galois orbit) is exactly the expected minimum:
  $$\dim_K \{F \in S_{n,d}(K) \mid F(P) = 0\} = \max(m - r, 0)$$
• Previous Work: Asgarli, Ghioca, and Reichstein [2] proved this result for fields $K$ with $|K| > 2$. They left the case $|K| = 2$ (specifically finite fields $\mathbb{F}_2$) as an open problem.
• Goal: The paper aims to resolve the $|K|=2$ case and provide a unified proof for all fields, including a generalization for arbitrary extension degrees $r$.

2. Methodology

The authors employ a multi-faceted approach combining algebraic geometry, combinatorics, and probabilistic counting arguments. The methodology is divided based on the nature of the base field $K$.

A. Infinite Fields

For infinite fields, the authors generalize the method used in [2].

• Inductive Construction: They construct points $P_1, \dots, P_r$ inductively to ensure linear independence of constraints.
• Algebraic Geometry Tools: They utilize the fact that the condition for "bad" points (where the dimension is larger than expected) is defined by the vanishing of certain polynomial functions (minors of a matrix). Since the base field is infinite, if these functions are not identically zero, a point exists where they do not vanish.

B. Finite Fields ($K = \mathbb{F}_q$)

The finite field case is significantly more challenging, particularly when $q=2$ and $r=m$. The authors develop three distinct methods to bound the number of "bad" points (points lying on too many hypersurfaces).

1. Method 1: Incidence Correspondence (Section 4)

   • Strategy: Count pairs $(P, H)$ where $P$ lies on a hypersurface $H$ in two ways: summing points per hypersurface and summing hypersurfaces per point.
   • Result: This provides a lower bound on the proportion of "good" points ($\mu$). It works well for most cases but fails to provide a tight enough bound when $q=2$ and $r=m$ due to the high density of points on hypersurfaces in small fields.

2. Method 2: Counting by Irreducible Components (Section 5)

   • Strategy: Instead of counting points on all hypersurfaces (many of which are reducible), the authors count points on irreducible hypersurfaces.
   • Refinement: They use the fact that reducible hypersurfaces are rare (proven via bounds on the number of reducible polynomials). By focusing on irreducible components, they obtain sharper bounds on the number of points.
   • Application: This method successfully handles the case $q \ge 3$ and the case $q=2$ when $d$ is sufficiently large relative to $n$.

3. Method 3: Inclusion-Exclusion Principle (Section 6)

   • Strategy: To handle the difficult case of $q=2$ with small $d$ and large $n$, the authors apply the inclusion-exclusion principle to the union of hypersurfaces.
   • Key Insight: They analyze the intersections of 2 and 3 hypersurfaces. The core difficulty is that intersections can be reducible. The authors adapt an argument by Poonen to show that "most" intersections of random hypersurfaces are geometrically irreducible and have small singular loci.
   • Execution: They bound the probability that an intersection is reducible or has a large singular locus. If the intersection is "good" (irreducible), standard point-counting bounds (Lang-Weil) apply. This allows them to prove the result for large $n$ even when $d$ is small.

C. Computational Verification

For the remaining "exceptional" cases where theoretical bounds are insufficient (e.g., very small $n, d$), the authors perform explicit computer searches using the Magma computer algebra system. They verify the existence of the required points for all remaining quadruples $(q, n, d, r)$.

3. Key Contributions and Results

Main Theorem (Theorem 1.2)

The authors prove that for any field $K$ and any separable extension $L/K$ of degree $r$, there exists a point $P \in \mathbb{P}^n(L)$ such that:
$$\dim_K \{F \in S_{n,d}(K) \mid F(P) = 0\} = \max\left(\binom{n+d}{n} - r, 0\right)$$

• Significance: This resolves the open question from [2] regarding the case $|K|=2$. It generalizes the Primitive Element Theorem to the context of hypersurfaces.

Applications to Linear Systems (Section 7)

The main theorem is applied to determine the maximum dimension of linear systems of hypersurfaces with specific properties:

• Irreducibility: They generalize a result by Asgarli et al. regarding linear systems where every member is irreducible.
• Large Irreducible Factors: They prove the existence of linear systems where every member has an irreducible factor of degree at least $i$.
• Reducedness: They determine the maximum dimension of a linear system where every member is a reduced hypersurface (square-free polynomial).
• Sharpness: In all cases, they prove these bounds are sharp (cannot be improved).

Conjecture on Point Proportions (Section 8)

The authors propose a conjecture (Conjecture 8.2) regarding the asymptotic proportion of points satisfying the condition as $n, d \to \infty$. They conjecture that for a fixed $q$, the proportion of points in $\mathbb{P}^n(\mathbb{F}_{q^m})$ not lying on any degree $d$ hypersurface over $\mathbb{F}_q$ converges to:
$$\prod_{i=1}^{\infty} (1 - q^{-i})$$
This suggests a probabilistic independence of monomial evaluations for random points.

4. Significance

1. Resolution of an Open Problem: The paper definitively settles the case of finite fields with 2 elements, which was the most difficult remaining case in the study of Galois orbits and hypersurfaces.
2. Unified Framework: It provides a comprehensive proof covering all fields (finite and infinite) and all extension degrees, unifying previous fragmented results.
3. Methodological Innovation: The combination of incidence counting, irreducible component analysis, and inclusion-exclusion with singular locus bounds represents a sophisticated toolkit for arithmetic geometry problems involving point-hypersurface incidences.
4. Optimization of Linear Systems: The results on linear systems (Theorem 1.4 and Corollary 7.3) provide precise limits on how large a family of hypersurfaces can be while maintaining geometric properties like irreducibility or reducedness over finite fields.
5. Computational Rigor: The inclusion of explicit verification for exceptional cases ensures the result is not just asymptotic but holds for all concrete parameters, bridging the gap between theoretical bounds and finite field arithmetic.

In summary, this paper advances the understanding of the interplay between field extensions and algebraic geometry, proving that "generic" behavior (points not lying on unexpected hypersurfaces) holds even in the most restrictive finite field settings.