The border rank of the $4 \times 4$ determinant tensor is twelve

Basic Information

• Paper ID: 2510.11051
• Title: The border rank of the $4 \times 4$ determinant tensor is twelve
• Authors: Jong In Han, Jeong-Hoon Ju, Yeongrak Kim
• Classification: math.AG (Algebraic Geometry), math.AC (Commutative Algebra)
• Publication Date: October 13, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.11051

Abstract

This paper proves that the border rank of the $4 \times 4$ determinant tensor over the complex field $\mathbb{C}$ is at least 12, using the fixed ideal theorem introduced by Buczyńska-Buczyński and the method of Conner-Harper-Landsberg. Combined with known upper bounds, this implies that the border rank is exactly 12.

Research Background and Motivation

Core Problem

The core problem addressed in this paper is determining the precise border rank of the $4 \times 4$ determinant tensor. The determinant tensor as a fourth-order tensor $\det_4 \in \mathbb{C}^4 \otimes \mathbb{C}^4 \otimes \mathbb{C}^4 \otimes \mathbb{C}^4$ is defined as: $\det_4 = \sum_{\sigma \in S_4} \text{sgn}(\sigma) e_{\sigma(1)} \otimes e_{\sigma(2)} \otimes e_{\sigma(3)} \otimes e_{\sigma(4)}$

Significance

1. Connection to Complexity Theory: The determinant polynomial appears in many important problems in complexity theory
2. Foundation of Geometric Complexity Theory: Studying the tensor rank and border rank of $\det_n$ and its Kronecker powers is foundational work in geometric complexity theory
3. Matrix Multiplication Exponent: Through Strassen's laser method, there exists a connection to the matrix multiplication exponent

Existing Results and Challenges

• Known: $R(\det_4) \leq 12$ (via explicit decomposition as a sum of 12 rank-1 tensors)
• Known: $R(\det_4) = 12$ and $11 \leq \overline{R}(\det_4)$ (via recursive Koszul flattening method)
• Open Problem: Determine whether the border rank $\overline{R}(\det_4)$ is 11 or 12

Core Contributions

1. Determination of Exact Border Rank: Proves that the border rank of the $4 \times 4$ determinant tensor is exactly 12
2. Methodological Contribution: Demonstrates the effective application of the fixed ideal theorem in studying border rank of higher-order tensors
3. Algorithmic Implementation: Provides explicit border rank determination criteria for fourth-order concise tensors
4. Theoretical Completion: Completes the rank complexity characterization of $\det_4$, namely $R(\det_4) = \overline{R}(\det_4) = 12$

Detailed Methodology

Task Definition

Given a tensor $T \in V_1 \otimes V_2 \otimes V_3 \otimes V_4$, the border rank $\overline{R}(T)$ is defined as the minimum integer $r$ such that $T$ can be expressed as the limit of a sum of $r$ rank-1 tensors.

Theoretical Framework

Border Dual Theory

For a tensor $T \in V_1 \otimes \cdots \otimes V_d$, its annihilator is defined as: $\text{Ann}(T) = \{\Theta \in \text{Sym}(V_1^*) \otimes \cdots \otimes \text{Sym}(V_d^*) | \Theta \lrcorner T = 0\}$

Fixed Ideal Theorem

Theorem 2.9: Let $T \in V_1 \otimes \cdots \otimes V_d$, and $H \subset G_T$ be a connected solvable group. If $\overline{R}(T) \leq r$, then there exists an ideal $I \subset \text{Sym}(V_1^*) \otimes \cdots \otimes \text{Sym}(V_d^*)$ corresponding to a border rank $r$ decomposition of $T$, and this ideal is $H$-invariant.

Algorithmic Strategy

Concise Tensor Property

$\det_4$ is a concise tensor, meaning all maps $T_{V_i}: V_i^* \to V_1 \otimes \cdots \otimes \hat{V_i} \otimes \cdots \otimes V_d$ are injective.

Testing Procedure

For a fourth-order tensor $T \in V_1 \otimes V_2 \otimes V_3 \otimes V_4$, verifying $\overline{R}(T) \leq r$ requires checking:

1. Inclusion Condition: $I \subset \text{Ann}(T)$
2. Codimension Condition: For all $(i_1,i_2,i_3,i_4)$ satisfying $r \leq \dim(S^{i_1}V_1^* \otimes \cdots \otimes S^{i_4}V_4^*)$, we have $\text{codim}(I_{i_1,i_2,i_3,i_4}) = r$
3. Multiplicative Closure: The multiplicative structure of the ideal
4. Invariance: Each $I_{i_1,i_2,i_3,i_4}$ is $B$-invariant

Experimental Setup

Group Action Configuration

• Choose $B$ as the Borel subgroup of $\text{SL}_4$ (upper triangular matrices)
• Decompose the Borel algebra: $\mathfrak{b} = \mathfrak{t} \oplus \mathfrak{n}$, where $\mathfrak{t}$ is the Cartan subalgebra and $\mathfrak{n}$ is the nilpotent part

Weight Analysis

Use the standard basis $\{v_1, v_2, v_3, v_4\}$ and weights $L_i \in \mathfrak{t}^*$ defined by $L_i(\text{diag}(a_1,a_2,a_3,a_4)) = a_i$.

Representation-Theoretic Decomposition

For $V^{\otimes 3} \cong S^{(1,2,3)}V \oplus S^{(1,2,3)}V \oplus S^{(1,3,2)}V \oplus S^{(1,2,3)}V$, where the last term is $\Lambda^3V$.

Experimental Results

Main Result

Theorem 1.1: The border rank of $\det_4$ is 12 over any subfield of $\mathbb{C}$.

Detailed Verification Process

Phase One: Low-Order Testing

All tests for $(2100), (1200), \ldots, (0012)$ pass because the corresponding map ranks are sufficiently small.

Phase Two: Candidate Space Determination

Through weight graph analysis, the unique $B$-invariant subspace is identified:

• $F_{1110}$ corresponds to $E_{1110} = \langle u_1(1,1,1), u_1(1,1,2), u_1(1,1,3)\rangle \oplus \langle u_2(1,1,2), u_2(1,1,3)\rangle \oplus \langle u_3(1,1,2), u_3(1,1,3)\rangle \oplus \Lambda^3V$

Critical Test: $(1111)$-Test

For the candidate 4-tuple $\{F_{1110}, F_{1101}, F_{1011}, F_{0111}\}$ passing all previous tests, the map: $(F_{1110} \otimes V_4^*) \oplus (F_{1101} \otimes V_3^*) \oplus (F_{1011} \otimes V_2^*) \oplus (F_{0111} \otimes V_1^*) \to V_1^* \otimes V_2^* \otimes V_3^* \otimes V_4^*$ has image dimension 246, strictly greater than $256 - 11 = 245$.

Conclusion

The $(1111)$-test fails, therefore $\overline{R}(\det_4) > 11$, and combined with the known upper bound, we obtain $\overline{R}(\det_4) = 12$.

Related Work

Tensor Rank Theory

• Tensor rank $R(T)$: minimum number of rank-1 tensors in a sum representation
• Border rank $\overline{R}(T)$: rank in the geometric closure sense, more suitable for algebraic geometry methods

Development of Dual Theory

• Classical dual theory: Sylvester's work, relating derivative ideals
• Multigraded generalization: Work of Gałązka et al.
• Border dual theory: Contributions of Buczyńska-Buczyński

Computational Methods

• Normal form lemma of Landsberg-Michałek
• Algorithmic framework of Conner-Harper-Landsberg
• Recursive Koszul flattening method

Conclusions and Discussion

Main Conclusions

1. The border rank of the $4 \times 4$ determinant tensor is exactly 12
2. For $n \leq 4$, we have $R(\det_n) = \overline{R}(\det_n)$
3. The effectiveness of the fixed ideal theorem in analyzing higher-order tensors is verified

Limitations

1. Computational Complexity: The method involves extensive representation-theoretic computations and computer-assisted verification
2. Generalization Difficulty: For cases $n \geq 5$, computational complexity increases dramatically
3. Theoretical Limitations: The tensor rank version of the fixed ideal theorem does not hold (as shown in Remark 3.2)

Future Directions

1. Open Question: Does $R(\det_n) = \overline{R}(\det_n)$ hold for all $n$?
2. Higher-Dimensional Cases: Determining the exact border rank of $\det_5$ (currently known: $27 \leq R(\det_5) \leq \overline{R}(\det_5) \leq 52$)
3. Method Improvement: Seeking more efficient methods for obtaining lower bounds on border rank

In-Depth Evaluation

Strengths

1. Theoretical Rigor: Employs profound tools from algebraic geometry and representation theory
2. Completeness: Thoroughly resolves the border rank problem for $\det_4$
3. Methodological Value: Demonstrates the practical application value of the fixed ideal theorem
4. Computational Verification: Provides detailed algorithmic implementation and verification procedures

Weaknesses

1. Computational Dependency: Critical steps require computer assistance, limiting method scalability
2. High Technical Threshold: Requires deep background in algebraic geometry and representation theory
3. Limited Generalizability: Direct application to higher-dimensional cases faces computational complexity barriers

Impact

1. Theoretical Contribution: Perfects the rank theory of small-dimensional determinant tensors
2. Methodological Influence: Provides a systematic algorithmic framework for border rank research
3. Geometric Complexity Theory: Provides important foundational results for related fields

Applicable Scenarios

1. Low-Dimensional Tensor Analysis: Suitable for symmetric tensors of small dimension
2. Theoretical Research: Provides tools for geometric complexity theory
3. Algorithm Development: Offers theoretical guidance for tensor decomposition algorithms

References

The paper cites 15 important references covering key works in tensor rank theory, dual theory, and computational methods. In particular, the border dual theory of Buczyńska-Buczyński and the computational method of Conner-Harper-Landsberg form the theoretical foundation of this paper.