Fedder type criteria for quasi-$F$-splitting I

Basic Information

• Paper ID: 2204.10076
• Title: Fedder type criteria for quasi-$F$-splitting I
• Authors: Tatsuro Kawakami, Teppei Takamatsu, Shou Yoshikawa
• Classification: math.AG (Algebraic Geometry), math.AC (Commutative Algebra), math.NT (Number Theory)
• Publication Date: April 2022 (arXiv v1), Latest version November 2025 (v3)
• Paper Link: https://arxiv.org/abs/2204.10076

Abstract

Yobuko recently introduced the concepts of quasi-$F$-splitting and quasi-$F$-split heights, which generalize and quantify the notion of Frobenius-splitting, and proved that for Calabi-Yau varieties, quasi-$F$-split heights coincide with Artin-Mazur heights. This paper establishes Fedder-type criteria for quasi-$F$-splitting of complete intersections, and in particular obtains simple formulas for computing Artin-Mazur heights of Calabi-Yau hypersurfaces. As one application, the authors prove the existence of Calabi-Yau varieties with arbitrarily large Artin-Mazur height over $\mathbb{F}_2$, and provide explicit defining equations of quartic K3 surfaces over $\mathbb{F}_3$ realizing all possible Artin-Mazur heights.

Research Background and Motivation

Core Problem to be Solved

The core problem addressed in this paper is: How to effectively determine the quasi-$F$-splitting property of positive characteristic algebraic varieties, particularly to compute their quasi-$F$-split heights.

Importance of the Problem

1. Theoretical Significance:
   • Frobenius splitting theory is a central tool in positive characteristic algebraic geometry; $F$-split varieties satisfy many desirable properties (such as Kodaira vanishing theorem)
   • Quasi-$F$-splitting is a natural generalization of $F$-splitting, refining the binary property (split or non-split) into a numerical invariant (height)
   • For Calabi-Yau varieties, quasi-$F$-split height equals Artin-Mazur height, which is a key invariant for stratifying moduli spaces
2. Computational Challenges:
   • Verifying quasi-$F$-splitting by definition requires constructing homomorphisms over Witt vector rings, involving complex ring structures
   • Direct computation of Artin-Mazur height is extremely difficult, especially for high-dimensional varieties

Limitations of Existing Methods

1. Classical Fedder Criterion: Only applies to $F$-splitting (height=1 case); for hypersurfaces $\{f=0\} \subset \mathbb{P}^n$, the criterion is $f^{p-1} \notin (x_0^p, \ldots, x_n^p)$
2. Definitional Approach: Requires constructing splitting homomorphisms over Witt vector rings $W_n(\mathcal{O}_X)$ of length $n$, with computational complexity increasing dramatically as $n$ grows
3. Lack of Systematic Tools: No unified algebraic criterion exists for computing arbitrary heights

Research Motivation

The authors' core motivation is to generalize Fedder's classical criterion to the quasi-$F$-splitting setting, providing a concise criterion requiring only the ring structure of length-2 Witt vectors, thereby making quasi-$F$-split heights (and Artin-Mazur heights of Calabi-Yau varieties) computable.

Core Contributions

The main contributions of this paper include:

1. Fedder-type Criterion (Theorem A, Theorem 4.11): For complete intersection local rings $R/I$ (where $I=(f_1,\ldots,f_m)$), provides a recursive algorithm for computing quasi-$F$-split height using only the ring structure of $W_2(R)$.
2. Simplified Criterion for Calabi-Yau Case (Theorem C, Theorem 5.8): For Calabi-Yau hypersurfaces (with $\deg f = N$), gives a particularly elegant formula: $\text{ht}(S/f) = \inf\{n \mid \theta^{n-1}(F_*^{n-1} f^{p-1}) \notin \mathfrak{m}^{[p]}\}$ where $\theta$ is an operator defined by $\Delta_1(f^{p-1})$.
3. Delta Formula (Theorem D, Theorem 3.20): Proves the crucial recursive formula $\Delta_n(f) \equiv f^{p^n} - p\Delta_1(f) \pmod{F(R)}$ This is the theoretical foundation for why the criterion requires only $W_2$ structure.
4. Explicit Constructions and Applications:
   • Constructs Calabi-Yau hypersurfaces of arbitrary even height $2h$ over $\mathbb{F}_2$ (Example 6.7)
   • Provides explicit equations for quartic K3 surfaces over $\mathbb{F}_3$ realizing all possible heights (1 to 10 and $\infty$) (Example 6.2)
   • Discovers a Calabi-Yau cubic hypersurface of height 60 over $\mathbb{F}_2$
5. Theoretical Tool Innovation:
   • Introduces pushout module $Q_{S,n}$ to simplify Witt vector module structures
   • Constructs splitting map $\sigma_n: Q_{R,n} \to F_*^{n-1}(F_*R/R)$
   • Establishes explicit description of homomorphism spaces $\text{Hom}_R(Q_{R,n}, R) \cong F_*R \oplus \bigoplus_{2\leq e\leq n} F_*^{e-1}\text{vK}$

Method Details

Task Definition

Input: Complete intersection variety $X = \text{Spec}(R/I)$ or $\text{Proj}(S/I)$ in positive characteristic $p$, where $I=(f_1,\ldots,f_m)$ is a regular sequence
Output: Quasi-$F$-split height $\text{ht}(X) \in \mathbb{Z}_{>0} \cup \{\infty\}$
Constraint: Use only basic ring-theoretic computations, avoiding direct treatment of high-length Witt vector rings

Core Concepts and Notation

Quasi-$F$-splitting Definition

A scheme $X$ is called $n$-quasi-$F$-split if there exists a $W_n\mathcal{O}_X$-module homomorphism $\phi: F_*W_n\mathcal{O}_X \to \mathcal{O}_X$ making the following diagram commute:

W_n O_X ---F---> F_* W_n O_X
   |                  |
R^{n-1}              φ
   |                  |
   v                  v
  O_X <------------- O_X

where $R$ is the restriction map and $F$ is the Frobenius map. The quasi-$F$-split height is defined as $\text{ht}(X) := \inf\{n \mid X \text{ is } n\text{-quasi-}F\text{-split}\}$

Key Technical Constructions

1. Pushout Module $Q_{S,n}$ (Definition 2.7)

To simplify Witt vector module structures, define the pushout:

W_n O_X ---F---> F_* W_n O_X
   |                  |
R^{n-1}               |
   |                  |
   v                  v
  O_X -----------> Q_{X,n}

Then $Q_{X,n} \cong F_*\overline{W}_n\mathcal{O}_X$ (where $\overline{W}_n = W_n/pW_n$), and $X$ is $n$-quasi-$F$-split if and only if $\mathcal{O}_X \to Q_{X,n}$ splits as an $\mathcal{O}_X$-module.

2. $p$-Monomials and $p$-Monomial Decomposition (Definition 3.4)

Let $R$ be an $F$-finite regular local ring, and choose a basis $\{F_*v_i\}$ such that $\{F_*(x_1^{i_1}\cdots x_N^{i_N}) \mid 0 \leq i_j \leq p-1\}$ is an $R$-basis of $F_*R$.

• An element $a \in R$ is called a $p$-monomial if $a = a_1^p v_i$ for some $a_1 \in R, i$
• A decomposition $a = a_1 + \cdots + a_r$ is called a $p$-monomial decomposition if each $a_i$ is a $p$-monomial

Key property: $p$-monomial decompositions are non-unique (e.g., $0 = v_i + (-1)^p v_i$), and this non-uniqueness plays a crucial role in constructing splittings.

3. $\Delta$ Map Family (Definition 3.7)

For $\alpha \in W(R)$, there uniquely exists $\alpha' \in \text{Im}(V)$ such that $\alpha - \alpha'$ is a sum of $p$-monomial Teichmüller lifts. Define $\Delta_W: W(R) \to W(R)/[F(R)]; \quad \alpha \mapsto V^{-1}\alpha'$

For $a \in R$, define $\Delta(a) := \Delta_W([a])$, with its 0-component denoted $\Delta_1(a)$. Inductively define $\Delta_n(a) := \sum_{r=1}^{n-1} \Delta_r \circ \delta_{n-r}(a) + \delta_n(a)$ where $\delta_i$ are defined by $\Delta(a) = (\delta_1(a), \delta_2(a), \ldots)$.

Technical Route of Main Theorems

Step 1: Constructing Splitting Maps (Section 3)

Proposition 3.10: There exists an $R$-module homomorphism $\sigma_n: Q_{R,n} \to F_*^{n-1}(F_*R/R)$ satisfying $\sigma_n(F_*(a_0, a_1, \ldots, a_{n-1})) = F_*^n(\Delta_{n-1}(a_0) + \Delta_{n-2}(a_1) + \cdots + a_{n-1})$ with $\sigma_n \circ V^{n-1} = \text{id}$.

Proof Strategy:

1. For $n=2$, directly define $\sigma_2(F_*(a,b)) = F_*^2(\delta_1(a) + b)$, verifying it is an $R$-module homomorphism using non-uniqueness of $p$-monomial decompositions
2. Inductively define $\sigma_n = \sigma_2 \circ F_*\sigma_{n-1}$
3. Use the exact sequence $0 \to F_*^{n-1}(F_*R/R) \to Q_{R,n} \to Q_{R,n-1} \to 0$ for induction

Step 2: Describing Homomorphism Spaces (Lemma 4.3)

Using the splitting $\sigma_n$, establish an isomorphism $\Psi_n: F_*R \oplus \bigoplus_{2\leq e\leq n} F_*^{e-1}\text{vK} \xrightarrow{\cong} \text{Hom}_R(Q_{R,n}, R)$ where vK$= \ker(u)$, with $u$ being a dual basis element. For $\psi_{(g_1,\ldots,g_n)} := \Psi_n((F_*g_1, \ldots, F_*^n g_n))$, there is an explicit formula: $\psi_{(g_1,\ldots,g_n)}(F_*(V^{s-1}[a])) = \sum_{r=0}^{n-s} u_{r+s}(F_*^{r+s}(g_{r+s}\Delta_r(a)))$

Step 3: Delta Formula (Theorem 3.20)

Key Result: $\Delta_n(a) \equiv a^{p^n} - p\Delta_1(a) \pmod{F(R)}$

Proof Strategy:

1. Lift to mixed characteristic: Consider $A = W(k)[x_1,\ldots,x_N]$ (a lift of $R$)
2. Define $\phi$-monomials on $A$ (elements satisfying $\phi(x) = x^p$)
3. Use ghost components (which behave well only in mixed characteristic) to prove $\Delta_n(a) = \frac{a^{p^n} - (a^p - p\Delta_1(a))^{p^{n-1}}}{p^n}$
4. Modulo $p$, the right side becomes $a^{p^n} - p\Delta_1(a)$

This formula shows that all $\Delta_n$ can be computed from $\Delta_1$, hence only $W_2(R)$ structure is needed!

Step 4: Main Criterion (Theorem 4.11)

Theorem A (Simplified Version): Let $f = f_1 \cdots f_m$, and define $\theta: \text{vK} \to R; \quad F_*a \mapsto u(F_*(\Delta_1(f^{p-1})a))$ Recursively define ideal sequence: $I_1 := (I^{[p]}:I)$, $I_{n+1} := \theta(F_*I_n \cap \text{vK}) + I_1$. Then $\text{ht}(R/I) = \inf\{n \mid I_n \not\subseteq \mathfrak{m}^{[p]}\}$

Proof Strategy:

1. By Lemma 4.3, $R/I$ is $n$-quasi-$F$-split if and only if there exist $g_1,\ldots,g_n$ satisfying:
   • $g_1 \notin \mathfrak{m}^{[p]}$, $u(F_*g_i)=0$ ($i \geq 2$)
   • For all $x \in I$, $\sum_{r=0}^{n-s} u_r(F_*^r(g_{r+s}\Delta_r(x))) \in I^{[ps]}$
2. Using the Delta formula, conditions simplify to recursive relations involving $\theta$
3. Lemma 4.9 equivalently transforms conditions into inclusion relations of ideal sequence $\{I_n\}$
4. Inductively construct $h_2,\ldots,h_n$ such that $h_s - \theta(F_*h_{s+1}) \in I_1$ and $\theta(F_*h_2) \notin \mathfrak{m}^{[p]}$

Step 5: Calabi-Yau Case Simplification (Theorem 5.8)

Theorem C: If $\deg f = \mu = \sum \deg x_i$ (Calabi-Yau condition), define $f_n := f^{p-1} \Delta_1(f^{p-1})^{1+p+\cdots+p^{n-2}}$ then $\text{ht}(S/f) = \inf\{n \mid f_n \notin \mathfrak{m}^{[p^n]}\} = \inf\{n \mid \theta^{n-1}(F_*^{n-1} f^{p-1}k) \not\subseteq \mathfrak{m}^{[p]}\}$

Proof Key Points:

1. Use homogeneity: $\deg(\Delta_1(f^{p-1})) = p(p-1)\mu$, $\deg(f_l) = (p^l-1)\mu$
2. Lemma 5.7: For homogeneous elements $a$ of degree $(p^l-1)\mu$, $a \in \mathfrak{m}^{[p^l]} \Longleftrightarrow u_l(F_*^l(ak)) = 0$
3. In Calabi-Yau case, $I_l$ is generated by $u_{l-1}(F_*^{l-1}(f_l k))$
4. Inductively prove $u_{l-1}(F_*^{l-1}(f_l k)) \subseteq I_l$

Technical Innovation Points

1. Pushout Technique: Replace $W_n\mathcal{O}_X$ with $Q_{S,n}$, transforming Witt vector module problems into more tractable $\mathcal{O}_X$-module problems
2. Exploiting Non-uniqueness of Decomposition: The non-uniqueness of $p$-monomial decomposition (usually viewed as a defect) becomes key to constructing splittings
3. Mixed Characteristic Lifting: Prove Delta formula in mixed characteristic using ghost components' good properties
4. Recursive Algorithm Design: Reduce height-$n$ determination to $n$ iterations of $\theta$, each requiring only $W_2$ computations
5. Homogeneity Simplification: In graded settings, use degree analysis to significantly simplify criteria

Experimental Setup

Computational Tools

This paper uses the Macaulay 2 computer algebra system to implement the criterion's algorithm, with code publicly available on the second author's homepage (link).

Computational Objects

1. Quartic K3 surfaces ($\mathbb{P}^3$ quartic hypersurfaces)
2. Quintic Calabi-Yau threefolds ($\mathbb{P}^4$ quintic hypersurfaces)
3. Hypersurfaces in weighted projective spaces (e.g., weighted Delsarte curves)

Computational Procedure

For hypersurface $\{f=0\}$:

1. Compute homogeneous representative of $\Delta_1(f^{p-1})$
2. Define operator $\theta: F_*S \to S$
3. Iteratively compute $f_n = f^{p-1}\Delta_1(f^{p-1})^{1+p+\cdots+p^{n-2}}$
4. Check $f_n \in \mathfrak{m}^{[p^n]}$ (via dual basis element $u_n$)
5. Output minimal $n$ such that $f_n \notin \mathfrak{m}^{[p^n]}$

Experimental Results

Main Results

1. Complete Classification of K3 Surfaces over $\mathbb{F}_3$ (Example 6.2, Table 1)

The authors provide explicit equations for quartic K3 surfaces realizing all possible Artin-Mazur heights (1 to 10 and $\infty$):

Significance: This is the first explicit construction realizing all possible heights over $\mathbb{F}_3$ (previously only known over $\mathbb{F}_2$, see KS16).

2. Calabi-Yau Varieties of Arbitrarily Large Height (Example 6.7)

Construction: For any positive integer $h$, take $N = 2h+1$ and define $f = a^N + b^N + c^N + x_1^N + \cdots + x_{N-3}^N + (b+c)g$ where $g = c^2x_1\cdots x_{N-3} + x_1^4x_2^2\cdots x_{N-3}^2 + \cdots + x_1^{2^{h-1}}x_2^{2^{h-1}}$

Result: $\text{ht}(\text{Proj}(S/f)) = 2h$

Proof Strategy:

1. Verify smoothness via partial derivative calculations
2. Lemma 6.6: If monomial decomposition $\alpha = M_1 + \cdots + M_n$ satisfies
   • $\min\{\text{ht}_f(M_i)\} = h$
   • $\#\{i \mid \text{ht}_f(M_i) = h\}$ is odd (over $\mathbb{F}_2$)
   
   then $\text{ht}_f(\alpha) = h$
3. Compute monomial heights term-by-term, verifying exactly 3 terms have height $2h$

Corollary: Over $\mathbb{F}_2$, Artin-Mazur heights of Calabi-Yau varieties can be arbitrarily large (answering an open problem in the field).

3. High-Height Calabi-Yau Threefold (Example 6.3)

The quintic hypersurface $x^5 + y^5 + z^5 + w^5 + u^5 + xz^3w + yzw^3 + x^2zu^2 + y^2z^2w + xy^2wu + yzwu^2 = 0$ has $\text{ht} = 60$ over $\mathbb{F}_2$.

Significance: This is the largest known height for smooth quintic Calabi-Yau threefolds (theoretical bound is 102).

4. Weighted Delsarte Surfaces (Example 6.1)

The hypersurface in weighted projective space $\mathbb{P}(1,2,3,8)$$x_0^{80}x_1 + x_1^{61}x_2 + x_2^{32} + x_3^2x_0 = 0$ has $\text{ht} = \infty$ over $\mathbb{F}_3$ (not quasi-$F$-split).

Verifying via minimal resolution that its K3 surface also has $\text{ht} = \infty$, recovering Goto's Got04 computation.

Experimental Findings

1. Height Parity: Calabi-Yau hypersurfaces constructed over $\mathbb{F}_2$ all have even heights, possibly reflecting characteristic 2 specificity.
2. Smoothness vs Height: Smoothness does not guarantee finite height (e.g., Fermat quartic $x^4+y^4+z^4+w^4$ has $\text{ht}=\infty$ over $\mathbb{F}_3$).
3. Computational Feasibility: Using Theorem C, even height-60 threefolds are computable (virtually impossible with traditional methods).
4. Height Jumps: In continuous deformation families, height can jump discontinuously (see Example 4.18 comparing $g$ and $gw$).

Related Work

Frobenius Splitting Theory

1. Mehta-Ramanathan MR85: Introduced $F$-splitting concept, proved Kodaira vanishing for $F$-split varieties
2. Fedder Fed83: Proved algebraic criterion for hypersurface $F$-splitting: $f^{p-1} \notin (x_0^p,\ldots,x_n^p)$
3. Hara Har98a,Har98b: Classified two-dimensional $F$-regular and $F$-pure singularities

Quasi-$F$-splitting

1. Yobuko Yob19,Yob20:
   • Introduced quasi-$F$-splitting and quasi-$F$-split height
   • Proved quasi-$F$-split height = Artin-Mazur height for Calabi-Yau varieties
   • Proved quasi-$F$-split varieties lift to $W_2(k)$
2. Achinger-Zdanowicz AZ21: Serre-Tate theory for Calabi-Yau varieties
3. Nakkajima-Yobuko NY21: Kodaira vanishing for quasi-$F$-split varieties

Artin-Mazur Height

1. Artin-Mazur AM77: Defined formal groups and height
2. van der Geer-Katsura vdGK00,vdGK03:
   • Stratification of K3 surface moduli spaces
   • Cohomological characterization of heights
3. Computational Results:
   • Stienstra Sti87, Yui Yui99, Goto Got04: Heights of specific K3 surfaces
   • Kedlaya-Sutherland KS16: Complete census of quartic K3s over $\mathbb{F}_2$

Position of This Work

This paper provides the first systematic algebraic tool for computing quasi-$F$-split heights, bridging the gap between theory (Yobuko's definition) and computation (specific examples). Compared to Yobuko's local cohomology approach, this paper's criterion is more suitable for computer implementation.

Conclusions and Discussion

Main Conclusions

1. Theoretical Achievement: Establishes Fedder-type criterion for quasi-$F$-splitting of complete intersections, requiring only $W_2$ ring structure
2. Computational Formula: For Calabi-Yau hypersurfaces, provides explicit single-element criterion $f_n$
3. Existence Result: Proves existence of Calabi-Yau varieties with arbitrarily large Artin-Mazur height over $\mathbb{F}_2$
4. Explicit Constructions: Provides K3 surface equations over $\mathbb{F}_3$ for all possible heights

Limitations

1. Complete Intersection Restriction: Theorem 4.11 applies only to complete intersections; general case requires further work (see KTY)
2. Characteristic 2 Specificity: Construction in Example 6.7 yields only even heights over $\mathbb{F}_2$; other characteristics handled in KTY
3. Non-Calabi-Yau Case: Simplification in Theorem 5.8 doesn't apply; must handle ideal sequence $\{I_n\}$
4. Computational Complexity: While theoretically requiring only $W_2$, determining large heights still requires multiple $\theta$ iterations
5. Infinite Height Determination: Corollary 4.19 gives partial sufficient conditions; complete characterization remains open

Future Directions

Authors address in subsequent work KTY:

1. Rational Double Points (RDP): Compute quasi-$F$-split heights of all RDPs (completing Yobuko's unpublished work)
2. Fano Varieties:
   • Construct smooth non-quasi-$F$-split Fano $d$-folds ($d>2$)
   • Prove del Pezzo surfaces have height $\leq 2$
3. Adjoint Inversion:
   • Fails in general (Example 4.18)
   • Holds for Calabi-Yau fibrations (Theorem 5.8)
4. Fiber Products: If $X$ non-$F$-split and $Y$ non-$F$-split, then $X \times Y$ non-quasi-$F$-split
5. Generic Fiber: Quasi-$F$-splitting doesn't always descend to generic fiber (wild cone bundle example)

Related work by other authors:

• Tanaka-Witaszek-Yobuko TWY24: quasi-$F^e$-splitting, quasi-$F$-regularity
• Yoshikawa Yos25: Fedder-type criteria for these concepts
• Kawakami-Takamatsu-Tanaka-Witaszek-Yobuko KTT+22,KTT+24: Applications to birational geometry

In-Depth Evaluation

Strengths

1. Theoretical Depth:
   • Delta formula proof (Theorem 3.20) demonstrates high technical sophistication; mixed characteristic lifting and ghost component usage are remarkably elegant
   • Pushout construction $Q_{S,n}$ is innovative for handling Witt vector modules
   • Exploitation of non-unique decomposition reflects deep insight
2. Computational Effectiveness:
   • Criterion is completely algorithmic; Macaulay 2 implementation verifies practical utility
   • Requiring only $W_2$ structure dramatically reduces computational complexity
   • Calabi-Yau simplification (Theorem C) is particularly elegant
3. Broad Applicability:
   • Solves long-standing computational problem for Calabi-Yau Artin-Mazur heights
   • Provides multiple new explicit examples
   • Subsequent work KTY demonstrates powerful applications
4. Writing Quality:
   • Clear structure progressing from motivation to technical details
   • Abundant examples aid understanding
   • Technical roadmap (Section 1.2) explanation is very helpful

Weaknesses

1. Technical Barriers:
   • Witt vector ring complexity makes full understanding difficult for non-specialists
   • Non-uniqueness of $p$-monomial decomposition, while key, lacks intuitive clarity
   • Necessity of mixed characteristic lifting (why not work directly in characteristic $p$?) insufficiently explained
2. Completeness:
   • Complete characterization of infinite height remains unresolved (Corollary 4.19 gives only partial conditions)
   • Non-complete-intersection case requires additional work
   • Optimality: Do simpler criteria exist?
3. Example Limitations:
   • Example 6.7 construction only over $\mathbb{F}_2$; other characteristics deferred to KTY
   • Few examples of high-dimensional Calabi-Yau (dimension $\geq 4$)
   • Systematic discussion of singular cases lacking
4. Algorithm Analysis:
   • No theoretical complexity analysis provided
   • Upper bounds on heights for given varieties not discussed
   • Computational bottlenecks not identified

Impact

1. Field Contribution:
   • Provides new computational tool for positive characteristic algebraic geometry
   • Connects commutative algebra (Fedder criterion) with arithmetic geometry (Artin-Mazur height)
   • Advances quasi-$F$-splitting theory development
2. Practical Value:
   • Macaulay 2 code available for community use
   • Makes previously intractable invariants computable
   • Provides methodology for constructing varieties with specified properties
3. Reproducibility:
   • Code public; results verifiable
   • Example equations completely explicit
   • Algorithm steps clearly described
4. Subsequent Impact:
   • Multiple follow-up works already exist (KTY, TWY24, Yos25, Yob23)
   • Applications in birational geometry (KTT+22,KTT+24)
   • Potential impact on moduli space theory (K3 surface height stratification)

Applicable Scenarios

1. Direct Applications:
   • Computing Artin-Mazur heights of Calabi-Yau hypersurfaces
   • Determining quasi-$F$-splitting of complete intersections
   • Constructing varieties with specified heights
2. Theoretical Research:
   • Studying lifting properties of positive characteristic varieties
   • Moduli space stratification structures
   • Frobenius-related cohomological properties
3. Computational Algebraic Geometry:
   • Developing more efficient algorithms
   • Database construction (e.g., K3 surface classification)
   • Computer-assisted theoretical discovery
4. Related Fields:
   • Formal group theory in arithmetic geometry
   • $F$-singularity theory in commutative algebra
   • Frobenius structures in representation theory

References

Key references include:

1. Yob19,Yob20 F. Yobuko's foundational work on quasi-$F$-splitting
2. Fed83 R. Fedder's classical Fedder criterion
3. AM77 M. Artin and B. Mazur's original formal group paper
4. vdGK03 G. van der Geer and T. Katsura's characterization of Artin-Mazur heights
5. BS22 B. Bhatt and P. Scholze's Prisms theory (provides modern perspective on $\Delta_1$)
6. KS16 K. S. Kedlaya and A. V. Sutherland's K3 surface census

Summary: This paper is an important contribution to positive characteristic algebraic geometry, successfully generalizing Fedder's classical ideas to quasi-$F$-splitting and providing powerful computational tools. The Delta formula proof demonstrates high technical sophistication, while abundant explicit examples verify theoretical practicality. Despite technical barriers and completeness gaps, this work establishes solid foundations for subsequent research and has already generated significant applications. For scholars studying positive characteristic varieties, Calabi-Yau varieties, or Frobenius-related properties, this is essential reading.