The 3 x 3 x 3 hyperdeterminant as a polynomial in the fundamental invariants for SL(3,C) x SL(3,C) x SL(3,C)

Basic Information

• Paper ID: 1310.3257
• Title: The 3 x 3 x 3 hyperdeterminant as a polynomial in the fundamental invariants for SL(3,C) x SL(3,C) x SL(3,C)
• Authors: Murray Bremner, Jiaxiong Hu, Luke Oeding
• Classification: math.AG (Algebraic Geometry), cs.SC (Symbolic Computation), math.RT (Representation Theory)
• Publication Date: February 17, 2014 (arXiv v2)
• Paper Link: https://arxiv.org/abs/1310.3257

Abstract

This paper briefly reviews prior work on invariant theory of 3×3×3 arrays, then recalls how to generate arrays of arbitrary size m₁×...×mₖ with zero hyperdeterminant. The main result provides an explicit formula expressing the 3×3×3 hyperdeterminant as a polynomial in the degree 6, 9, and 12 fundamental invariants under the action of the Lie group SL₃(C)×SL₃(C)×SL₃(C). The authors apply their computational results to Nurmiev's classification of canonical forms for 3×3×3 arrays.

Research Background and Motivation

Core Problem

The central problem addressed in this paper is determining the coefficients in the explicit polynomial expression of the 3×3×3 hyperdeterminant Δ₃₃₃ in terms of the fundamental invariants I₆, I₉, I₁₂.

Research Significance

1. Theoretical Importance: Invariant theory of 3×3×3 arrays is a classical problem in algebraic geometry and representation theory, tracing back to foundational work by Aronhold (1850) and Cayley (1845)
2. Computational Complexity: Hyperdeterminant computation is extremely complex, requiring manipulation of polynomials with enormous numbers of terms (e.g., I₁₂ has 209,061 terms)
3. Applied Value: Important applications in quantum computing, black hole physics, and multilinear algebra

Limitations of Existing Methods

• The Schläfli method for computing the 3×3×3 hyperdeterminant requires substantial memory
• Classical invariant theory methods, while theoretically sound, are computationally intractable
• Explicit coefficient expressions for fundamental invariants have been lacking

Research Motivation

Vinberg (1976) proved that the algebra of invariants is freely generated by I₆, I₉, I₁₂, but the specific coefficients in the hyperdeterminant expression remained unknown. This paper aims to determine these coefficients through computational algebra methods.

Core Contributions

1. Main Theorem: Provides an explicit formula for the 3×3×3 hyperdeterminant: $\Delta_{333} = I_3^6 I_9^2 - I_2^6 I_{12}^2 + 36 I_6 I_9^2 I_{12} + 108 I_9^4 - 32 I_{12}^3$
2. Computational Method: Develops an efficient computational approach based on modular arithmetic and rational reconstruction
3. Theoretical Application: Applies results to Nurmiev's canonical form classification, verifying invariant values on various families of canonical forms
4. Rank Analysis: Determines the vanishing properties of invariants on arrays of different ranks

Methodology Details

Problem Formulation

Given the general form of the 3×3×3 hyperdeterminant: $\Delta_{333} = aI_6^6 + bI_4^6I_{12} + cI_3^6I_9^2 + dI_2^6I_{12}^2 + eI_6I_9^2I_{12} + fI_9^4 + gI_{12}^3$

The objective is to determine the coefficients a, b, c, d, e, f, g.

Core Algorithm

1. Zero Hyperdeterminant Array Generation

Using Lemma 2.3, generate arrays with zero hyperdeterminant through multilinear coordinate transformations:

• Set μᵢ₁...ᵢₖ = 0 when k-1 indices equal 1
• Apply pseudo-random basis transformations to ensure generality

2. Modular Arithmetic Method

• Select prime p = 10007
• Generate 10 pseudo-random zero hyperdeterminant arrays
• Compute fundamental invariants modulo p
• Establish a system of linear equations for the coefficients

3. Rational Reconstruction

Use Maple's iratrecon procedure to reconstruct modular p results as rational number coefficients.

Technical Innovations

1. Efficient Computational Strategy: Avoids direct hyperdeterminant computation by employing linear algebra methods
2. Modular Arithmetic Optimization: Uses modular arithmetic to circumvent complexity of large integer calculations
3. Verification Mechanism: Double verification through rational arithmetic and integer computation

Experimental Setup

Computational Environment

• Maple computer algebra system
• Modular arithmetic using prime p = 10007
• Integer computation verification using 343 non-zero arrays from {0,1}³

Data Generation

• Pseudo-random 3×3×3 arrays satisfying the zero hyperdeterminant condition
• Ensures all fundamental invariant values are non-zero to avoid degenerate cases

Verification Methods

1. Modular arithmetic computation of coefficients
2. Rational reconstruction verification
3. Independent verification through integer arithmetic

Experimental Results

Main Results

Explicit formula obtained through computation: $\Delta_{333} = I_3^6I_9^2 - I_2^6I_{12}^2 + 36I_6I_9^2I_{12} + 108I_9^4 - 32I_{12}^3$

Coefficient Verification

The solution space of the linear system has dimension 1, uniquely determining the coefficient vector: $[a, b, c, d, e, f, g] = [0, 0, -\frac{1}{32}, \frac{1}{32}, -\frac{9}{8}, -\frac{27}{8}, 1]$

Nurmiev Canonical Form Application

Verified invariant values on five canonical form families:

• First family: All invariants potentially non-zero
• Second family: Δ = 0
• Third family: I₉ = I₁₂ = Δ = 0
• Fourth family: Δ = 0
• Fifth family: All invariants vanish

Rank Analysis Results

Related Work

Historical Development

• Classical Period: Foundational work by Aronhold (1850) and Cayley (1845)
• Modern Development: Vinberg's (1976) Lie group methods, Gelfand et al.'s (1992) hyperdeterminant theory
• Computational Aspects: Strassen (1983), Ottaviani (2007) determinant formulas

Relationship to This Work

This paper builds upon Vinberg's free generation result, resolving the long-standing problem of computing explicit coefficients.

Conclusions and Discussion

Main Conclusions

1. First explicit polynomial formula for the 3×3×3 hyperdeterminant in terms of fundamental invariants
2. Verification of computational results for consistency with Nurmiev's classification
3. Completion of the theory of invariant vanishing properties on arrays of different ranks

Limitations

1. Methods primarily applicable to 3×3×3 case; generalization to higher dimensions requires additional work
2. Computational complexity remains high, particularly for larger arrays
3. Theoretical analysis primarily based on numerical computation, lacking purely algebraic proof

Future Directions

1. Generalization to hyperdeterminants of higher-dimensional arrays
2. Development of more efficient computational algorithms
3. Exploration of applications in quantum information and physics

In-Depth Evaluation

Strengths

1. Computational Breakthrough: Resolves a long-standing computational problem
2. Methodological Innovation: Cleverly combines modular arithmetic and rational reconstruction
3. Comprehensive Verification: Multiple cross-verification methods ensure result reliability
4. Theoretical Application: Successfully applies results to canonical form classification theory

Weaknesses

1. Computational Dependence: Relies primarily on numerical computation rather than pure algebraic methods
2. Generalization Difficulty: Methods difficult to directly extend to more general cases
3. Theoretical Depth: Lacks deeper theoretical explanation of coefficient structure

Impact

1. Theoretical Contribution: Provides important concrete results for invariant theory
2. Computational Value: Furnishes foundational results for numerical computation in related fields
3. Application Potential: Broad application prospects in quantum information and algebraic geometry

Applicable Scenarios

• Theoretical research in multilinear algebra
• Entanglement measures in quantum information
• Invariant computation in algebraic geometry
• Tensor decomposition and rank computation problems

References

The paper contains 32 references spanning from 19th-century classical invariant theory to modern computational algebraic geometry, providing readers with a complete historical context and theoretical background.