Secondary invariants and non-perturbative states

✨

This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to describe a complex dance performance. You have a troupe of $N$ dancers (the "matrices"), and they are constantly spinning, swapping places, and moving in perfect synchronization.

In physics, we often want to ignore the specific choreography of who is standing where and focus only on the shape of the dance itself. This is called finding the "gauge-invariant" description. It's like describing the dance by saying, "The group formed a circle," rather than "Alice is at the north, Bob is at the south."

This paper tackles a deep mystery: What happens when the number of dancers ( $N$ ) is small and finite, rather than infinite?

The Big Idea: The "Seed" and the "Tower"

The authors propose a new way to think about the states of a quantum system (the possible dances). They use a mathematical tool called the Hironaka decomposition, which splits the description of the system into two distinct parts:

Primary Invariants (The Perturbative "Tower"):
Think of these as the standard, predictable moves. If you have a basic dance move, you can repeat it, do it twice, or do it three times. These are like building blocks. In physics, these are the "perturbative" states—the easy ones we can calculate using standard methods. They form a continuous "tower" of possibilities.
Secondary Invariants (The Non-Perturbative "Seeds"):
This is the paper's big revelation. These are special, unique "seed" states. You can't build them by just repeating the basic moves. They are distinct, isolated starting points.
- The Analogy: Imagine a forest. The "Primary" invariants are the trees growing from the ground (continuous, predictable). The "Secondary" invariants are the seeds buried deep underground. You can't see them by just looking at the trees. They represent a different "sector" or "branch" of reality.

The paper argues that the full quantum universe isn't just one big forest of trees. It's a collection of different forests, each growing from a different "seed" (secondary invariant). The "Primary" moves just build the trees within that specific forest.

The Experiment: Simplifying the Dance

To prove this isn't just abstract math, the authors looked at very simple "zero-dimensional" models. Imagine the dancers aren't moving in time, but are just frozen in a single pose. They studied groups of 2, 3, and 4 matrices (dancers) and asked: "If we describe the dance using only the shape (invariants), how many different 'forests' (branches) do we find?"

Here is what they found:

2 Matrices: There is only one forest. Everything is connected. (Trivial).
3 Matrices: Suddenly, there are two forests. They look identical from the outside, but they are mirror images of each other (like left and right hands). To describe the full system, you must sum over both.
4 Matrices: The complexity explodes. There are eight distinct forests.
N Particles: If you have $N$ particles swapping places, there are $N!$ (N factorial) forests. For just 10 particles, that's 3.6 million different branches!

The "Branch" Metaphor

The authors show that when you rewrite the math to focus only on the "shape" of the dance, the calculation doesn't just become a single integral (a single path). It becomes a sum over many paths.

The Primary Variables: These are the smooth, continuous variables you integrate over (like the height of the trees).
The Secondary Variables: These act like discrete switches. They tell you which branch of the forest you are currently calculating.

In standard physics (perturbation theory), we usually only look at one branch. We assume the "seed" is just the empty vacuum. This paper argues that at finite $N$ , you are missing the other branches. These other branches correspond to non-perturbative states—things like Black Holes in the universe of gravity.

Why This Matters: The Black Hole Connection

In the theory of Black Holes, we know they have a huge amount of "entropy" (disorder). This entropy comes from the number of microscopic ways a black hole can be arranged.

The math of these "Secondary Invariants" (the seeds) grows exponentially with the size of the system ( $e^{N^2}$ ).
This matches the growth of Black Hole entropy perfectly.

The authors suggest that Black Holes are the "Secondary Invariants." They are the special, non-perturbative "seeds" that sit outside the normal perturbative description. The "Primary" invariants are just the ripples on the surface of the black hole, while the "Secondary" invariants are the black hole itself.

The "Saddle Point" Discovery

Finally, the authors did something clever. They showed that these different "branches" aren't just mathematical tricks. They appear naturally as saddle points (critical peaks and valleys) in a new kind of mathematical landscape.

Imagine you are hiking in a mountain range. Standard physics only looks at the main valley.
This paper shows that there are hidden mountain passes (saddles) that connect to other valleys.
Even if you can't walk there directly in the real world, these "passes" are mathematically real and necessary to get the correct answer.

Summary in One Sentence

This paper reveals that the quantum universe is not a single, smooth landscape, but a complex web of multiple, distinct "branches" (sectors); the "primary" variables describe the smooth terrain within a branch, while the "secondary" variables are the hidden keys that unlock the existence of entirely new, non-perturbative worlds (like Black Holes) that standard physics usually misses.

1. Problem Statement

The paper addresses the structure of the algebra of gauge-invariant operators in matrix models and gauge theories at finite $N$ .

Infinite $N$ Limit: The operator algebra is freely generated by multi-trace operators, corresponding to a Fock space of perturbative excitations (supergravity states in holographic duals).
Finite $N$ Limit: Trace relations impose constraints, making the invariant ring non-free. The algebra becomes more intricate, and the Hilbert space structure is not simply a free Fock space.
The Core Question: How can the finite- $N$ structure of the invariant ring be physically interpreted? Specifically, the authors investigate the Hironaka decomposition, which expresses the invariant ring $R$ as a free module over a polynomial ring $P$ generated by primary invariants, with a basis of secondary invariants.
$R = \bigoplus_{\alpha=0}^{N_S-1} P \cdot s_\alpha$
The authors propose a physical interpretation: Primary invariants correspond to perturbative degrees of freedom (oscillators), while secondary invariants correspond to distinguished non-perturbative states or sectors (backgrounds) upon which perturbative towers are built.

2. Methodology

The authors test this interpretation using zero-dimensional multi-matrix integrals (matrix models) with $N=2$ and varying numbers of matrices ( $d=2, 3, 4$ ), as well as a system of $N$ bosons with $S_N$ symmetry.

Change of Variables: They explicitly perform the change of variables from matrix entries to invariant variables (traces and Gram matrix elements).
Geometric Analysis: They analyze the geometry of the quotient space (the space of orbits under the gauge group $U(N)$ ). They demonstrate that the map from the full configuration space to the space of primary invariants is not one-to-one but forms a finite algebraic cover.
Branch Counting: They count the number of "sheets" (branches) in this algebraic cover and compare it to the number of secondary invariants predicted by the Hironaka decomposition.
Saddle Point Analysis: They construct effective actions in the space of invariant variables. They show that the algebraic branch structure (the discrete fiber) arises naturally as the critical locus (saddles) of these invariant-space actions.
Verification: They verify the validity of their change of variables by reproducing exact Gaussian correlators in the original matrix variables using the reduced invariant integrals.

3. Key Contributions and Results

A. Explicit Change of Variables for $N=2$ Matrix Models

The authors derive the reduced measures and integration domains for $d=2, 3, 4$ Hermitian matrices.

Two Matrices ( $d=2$ ):
- Invariants: 5 primary invariants ( $p_1, \dots, p_5$ ) and 1 trivial secondary ( $s_0=1$ ).
- Geometry: The quotient is a single branch over the primary space. The integration domain is defined by the positivity of the Gram matrix of the associated vectors.
- Result: The integral reduces to a standard integral over the 5 primary invariants.
Three Matrices ( $d=3$ ):
- Invariants: 9 primary invariants and 2 secondary invariants ( $s_0=1$ , $s_1 = \text{Tr}(M_1 M_2 M_3)$ ).
- Geometry: The quotient forms a two-sheeted cover over the primary space. The sheets are distinguished by the sign of the oriented volume $\tau = \vec{m}_1 \cdot (\vec{m}_2 \times \vec{m}_3)$ , which relates to the imaginary part of the cubic trace.
- Result: The integral becomes a sum over two branches ( $\epsilon = \pm 1$ ). The measure includes a factor $1/\sqrt{\det G}$ .
Four Matrices ( $d=4$ ):
- Invariants: 13 primary invariants and 8 secondary invariants (1 trivial, 7 non-trivial).
- Geometry: The complexified quotient is an eight-sheeted cover.
  - Four sheets arise from a quartic equation $\Phi(p, \Delta^{(4)}) = 0$ (where $\Delta^{(4)}$ is a sum of diagonal Gram entries).
  - A further doubling arises from the sign of cubic invariants (oriented volumes).
- Result: The integral decomposes into a sum over 8 branches. The real Hermitian contour selects a subset of these branches based on positivity constraints, but the full algebraic structure is an 8-sheeted cover.

B. $S_N$ Symmetry Example (N Bosons in 2D)

System: $N$ bosons with diagonal $S_N$ gauge symmetry.
Invariants: $2N$ primary invariants (power sums) and $N!$ secondary invariants.
Result: The configuration space modulo gauge symmetry is an $N!$ -sheeted cover of the primary space. The secondary invariants encode the permutation $\sigma \in S_N$ that pairs the $x$ -roots and $y$ -roots. The integral explicitly reorganizes into a sum over $N!$ branches.

C. Saddle Point Interpretation

The authors construct effective actions in the space of invariant variables (introducing Lagrange multipliers to enforce algebraic constraints like $\det G = 0$ or $\Phi(p, \Delta)=0$ ).

Finding: The critical points (saddles) of these effective actions correspond exactly to the algebraic branches identified in the change of variables.
Significance: This demonstrates that the branch structure is not merely a formal algebraic artifact but arises from a genuine variational problem. The secondary invariants label the discrete sectors (saddles) of the theory.

4. Physical Significance and Interpretation

Perturbative vs. Non-Perturbative:
- Primary Invariants: Act as continuous variables governing fluctuations within a specific sector. They correspond to the perturbative Fock space excitations.
- Secondary Invariants: Encode the discrete data distinguishing different algebraic sectors (sheets) lying above the same primary data. They correspond to non-perturbative states or backgrounds.
Holographic Connection:
- In holographic duals (AdS/CFT), black hole microstates are expected to dominate the Hilbert space at finite $N$ .
- The number of secondary invariants grows as $\sim e^{cN^2}$ , matching the Bekenstein-Hawking entropy scaling ( $S_{BH} \sim N^2$ ).
- The authors argue that these secondary invariants naturally identify the non-perturbative sectors (black hole microstates), while the primaries describe the perturbative excitations (gravitons) on top of these backgrounds.
Semiclassical Picture: The decomposition of the path integral into a sum over branches mirrors the semiclassical expansion around different saddle points. The secondary invariants label these distinct saddles.

5. Conclusion

The paper provides strong evidence that the Hironaka decomposition of the invariant ring has a direct physical realization in matrix models. The algebraic structure of the quotient space (a finite cover over primary variables) naturally separates the theory into:

A continuous base (primaries) describing perturbative fluctuations.
A discrete fiber (secondaries) describing non-perturbative sectors.

This framework offers a concrete algebraic mechanism for understanding how finite- $N$ effects organize the Hilbert space into non-perturbative sectors, potentially providing a new language for describing black hole microstates and the transition from perturbative to non-perturbative physics in gauge theories.