⚛️ high-energy theory

Black Holes and Random Variables

This paper proposes that black hole microstate counts and their conformal field theory duals exhibit extreme value statistics characteristic of Gaussian log-correlated fields, thereby establishing a quantitative limit on the resolution of the semiclassical AdS gravitational path integral through a holographic formulation of the Fyodorov-Hiary-Keating conjecture.

Original authors: Eric Perlmutter

Published 2026-07-03

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Eric Perlmutter

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ 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

The Big Picture: Black Holes as Chaotic Crowds

Imagine a black hole not as a smooth, empty void, but as a massive, chaotic crowd of invisible particles (called "microstates"). In classical physics, we treat this crowd like a smooth fluid. We can count the crowd and say, "There are about a trillion people here." This is the "average" count.

However, the author of this paper argues that if you look at this crowd with a super-microscope (quantum mechanics), the count isn't smooth at all. It's jittery, erratic, and full of tiny, random bumps and dips. The paper proposes a new mathematical rule to describe exactly how "jittery" these counts can get.

The Core Idea: The "FHK" Rule

The paper borrows a famous idea from mathematics called the FHK Conjecture (named after Fyodorov, Hiary, and Keating).

The Original Context: Mathematicians originally used this rule to study the Riemann Zeta function (a famous equation in number theory) and random matrices (like a giant grid of random numbers). They found that the highest peaks and lowest valleys in these systems follow a very specific, predictable pattern of randomness.
The New Application: Perlmutter suggests that Black Holes and the Conformal Field Theories (CFTs) that describe them (via the AdS/CFT correspondence) follow this exact same pattern.

Think of it like this: If you listen to the static noise of a radio, you might expect it to be random. But if you look closely at the loudest spikes in that noise, they aren't just random; they follow a specific "law of the extreme." This paper says black hole microstates obey that same law.

The "Erratic" Fluctuations

The paper focuses on interval counts. Imagine you have a long ruler measuring the energy of a black hole. You pick a small segment of the ruler (an interval) and count how many microstates fit inside.

The Average: You can predict the average number of microstates in that segment very well using standard gravity equations. This is like knowing the average temperature of a room.
The Fluctuation: But the actual number will wiggle around that average. The paper claims these wiggles aren't just random noise; they are "extreme value statistics."

The Analogy of the Mountain Range:
Imagine the energy levels of a black hole are like a mountain range.

The average is the general height of the terrain.
The fluctuations are the jagged peaks and deep valleys.
The paper says that the highest peaks in this range are taller than you would expect from simple randomness. They are "log-correlated," meaning the height of one peak is subtly connected to the height of its neighbors, creating a complex, bumpy landscape.

The "Random Variable" (The Mystery Box)

The most important finding is that these extreme wiggles are governed by a specific random variable (let's call it $Y$ ).

What it is: It's a mathematical "mystery box" that determines how much the actual count of black hole states deviates from the smooth average.
The Tail: The paper describes the "tail" of this variable. In plain English, this means: "How likely is it to see a huge deviation?" The paper says these huge deviations happen more often than standard randomness would predict, following a specific formula ( $y e^{-2y}$ ).
Why it matters: This variable $Y$ is "order one," meaning it doesn't get huge as the black hole gets bigger. It stays small but significant.

The Limit of Our Knowledge (The "Precision Limit")

This is the most practical takeaway for physicists.

In the world of Semiclassical Gravity (our current best theory of how gravity works on large scales), we calculate things using a "path integral." Think of this as a recipe that sums up all possible ways a black hole could exist to get an answer.

The Problem: This recipe gives us a smooth, perfect answer.
The Reality: The actual quantum answer is jagged and noisy.
The Conclusion: The paper proves that the "noise" (the random variable $Y$ ) creates a hard limit on how precise our smooth recipe can ever be.

The Metaphor:
Imagine you are trying to measure the weight of a feather using a scale that is accurate to the nearest gram. You get a reading of 1 gram. But the paper says, "Actually, the feather is vibrating so wildly that the true weight fluctuates by a tiny, unmeasurable amount."
The paper quantifies this: The error in our semiclassical gravity calculations is roughly $e^{-S_0}$ (where $S_0$ is the black hole's entropy, a measure of its size).

For a huge black hole, this error is incredibly tiny, but it is not zero.
It means our current "smooth" theory of gravity cannot resolve the tiny, jagged details of the quantum world. There is a "fuzziness" that is fundamental and unavoidable.

Summary of Claims

Black Holes are Random Matrices: The spectrum of black hole microstates behaves like a random matrix, specifically one with "log-correlated" statistics (where nearby values are linked).
Extreme Statistics: The highest and lowest counts of these microstates follow the FHK Conjecture. They are not just random; they follow a specific, proven mathematical law for extreme values.
The "Glassy" Landscape: The energy landscape of a black hole is "glassy." It's not a smooth hill; it's a complex terrain of peaks and valleys that is hard to navigate, similar to how glass is a disordered solid.
The Precision Limit: Because of these random fluctuations, the semiclassical description of gravity (which assumes a smooth spacetime) has a built-in "blur." It cannot predict the exact number of microstates in a small energy window; it can only predict the average, with a specific, irreducible error margin.

What the Paper Does NOT Say

It does not propose a new way to build a black hole.
It does not claim we can use this to communicate faster than light.
It does not say this applies to every type of black hole in every universe, but specifically to large black holes in the context of AdS/CFT (a specific theoretical framework).
It does not claim to have "solved" quantum gravity, but rather to have placed a precise "speed limit" on how well our current theories can describe the quantum details.

In short: Black holes are smoother than they look, but not smooth enough to be perfectly predictable. There is a tiny, random "static" in the universe's background noise that our current math cannot fully eliminate.

Technical Summary: Black Holes and Random Variables

Problem Statement
The paper addresses the statistical nature of high-energy spectra in quantum gravity and conformal field theory (CFT), specifically focusing on the microstate counts of large AdS black holes. While the coarse-grained asymptotic density of states is well-understood via the AdS/CFT correspondence and conformal bootstrap methods (e.g., Cardy-like formulas), the fine-grained structure of interval counts—specifically the fluctuations of the number of primary operators within a mesoscopic energy window—remains less established. The authors seek to formulate a precise conjecture regarding the extreme value statistics of these counts, drawing an analogy to the Fyodorov-Hiary-Keating (FHK) conjecture from random matrix theory (RMT) and number theory. The central problem is to determine whether the erratic, order-one fluctuations in black hole microstate counts are governed by the universal statistics of log-correlated fields, and what this implies for the precision limits of the semiclassical gravitational path integral.

Methodology
The authors employ a multi-step methodology bridging random matrix theory, analytic number theory, and holographic duality:

Review of FHK Conjectures: The paper begins by reviewing the FHK conjectures, which posit that the extreme values of log-characteristic polynomials of random matrices (and the Riemann zeta function) follow a specific universal law. This law involves a leading order term scaling as $\log N$ , a subleading correction of $-\frac{3}{4}\log \log N$ , and a limiting random variable $Y_\infty$ with a prescribed tail distribution $P(Y_\infty > y) \sim y e^{-2y}$ .
Imaginary FHK and Counting Statistics: The authors adapt these conjectures to the "imaginary" parts of log-determinants. They demonstrate that the imaginary part of the log-determinant, $\Phi(z) = \Im \log P_N(z)$ , is directly related to the interval count of eigenvalues (or zeros). By analyzing the supremum and infimum of this field, they derive a conjecture for the fluctuations of interval counts, $\delta N_I = N_I - \bar{N}_I$ , where $N_I$ is the actual count and $\bar{N}_I$ is the mean.
AdS/CFT Formulation: The authors map this framework to AdS/CFT. They identify the interval counts of high-dimension primary operators in a CFT (dual to black hole microstates) with the eigenvalue counts in RMT. They define the counting fluctuation field $\delta N^{(J)}_I$ for operators with fixed spin $J$ and propose that these fluctuations obey the FHK law as the conformal dimension $\Delta \to \infty$ .
Gaussian Random Models: The paper utilizes a random Fourier model to describe the counting fluctuations. This model arises naturally from the diagonalization of a mesoscopic central limit theorem (CLT) with logarithmic variance, effectively treating the spectral density fluctuations as a log-correlated Gaussian field.
Large $N$ Analysis: The authors analyze the conjecture in the large $N$ limit (where the bulk theory is semiclassical AdS gravity). They relate the mean interval count to the smooth semiclassical gravitational path integral result and the erratic fluctuations to non-perturbative corrections.

Key Contributions and Results

Formulation of the AdS/CFT FHK Conjecture: The paper proposes that for high-energy primary operators in a CFT, the supremum of the counting fluctuations over mesoscopic intervals follows the law:
$\sup_{I \subseteq I_\ell} (\sigma \delta N^{(J)}_I) = \frac{2}{\pi} \sqrt{\frac{2}{\beta}} \left( \log \ell - \frac{3}{4} \log \log \ell + Y^{(\sigma|J)}_\ell \right)$
where $\ell$ is the unfolded length of the interval, $\beta$ encodes the symmetry class (GOE, GUE, or GSE), and $Y^{(\sigma|J)}_\ell$ is an order-one random variable converging to a limiting variable with the tail $P(Y_\infty > y) \sim y e^{-2y}$ .
Bootstrap Bounds: The authors interpret the leading-order term of this conjecture as a rigorous statistical bound for CFTs. They argue that this result provides an optimal bound for interval counts in $d$ -dimensional CFTs, strengthening existing Tauberian bounds (which are currently limited to $d=2$ or specific rational CFTs) by incorporating spectral statistics.
Precision Limit of Semiclassical Gravity: A primary result is the identification of a quantitative limit on the semiclassical gravitational path integral. The mean interval count corresponds to the smooth semiclassical result ( $N^{(grav)}_I \sim e^{S_0}$ ), while the erratic fluctuations are of order $O(1)$ in the unfolded variable, corresponding to a suppression of $O(e^{-S_0})$ relative to the mean. This implies that the Euclidean semiclassical path integral cannot resolve spectral details finer than $e^{-S_0}$ due to irreducible stochastic fluctuations.
Saddle Completeness: The equality nature of the conjecture (as opposed to a one-sided bound) suggests a "saddle completeness relation." It implies that if there are exponentially large contributions to the density of states (terms scaling as $e^{\alpha S_0}$ with $\alpha > 0$ ), they must be fully accounted for by the smooth semiclassical saddle sum; otherwise, the extremal fluctuations would violate the conjectured bound.
Decorrelation in $N$ : The paper connects the erratic $N$ -dependence of holographic CFT data to the decorrelation properties of the Gaussian random models. It suggests that the "erraticness" observed in wormhole amplitudes corresponds to a rapid decorrelation of the random Fourier coefficients as the parameter $N$ changes.

Significance
The paper claims to seed future explorations of extreme value theory and log-correlated statistics within high-energy physics. Its significance lies in:

Unifying Framework: It unifies the statistical mechanics of disordered landscapes (via the FHK conjecture) with the microstate counting of black holes, suggesting that black hole spectra lie in the universality class of Gaussian log-correlated fields.
Refining Semiclassical Gravity: It provides a precise, quantitative proposal for the limitations of the semiclassical approximation in the microcanonical ensemble, moving beyond heuristic arguments to specific statistical bounds.
Bootstrap Advancement: It offers a new perspective on the conformal bootstrap, framing the problem of bounding spectral data as an extreme value theory problem, potentially leading to sharper bounds in dimensions $d > 2$ .
Universality: It posits that the extreme value statistics of deterministic chaotic systems (like CFTs) should mirror those of random matrices, extending the known universality of local spectral correlations to the realm of global extremal fluctuations.

The authors remain modest regarding the proof status, noting that while the core of the FHK conjecture for random matrices is substantially proven, the application to specific CFTs and the precise form of the limiting random variable $Y_\infty$ in the gravitational context remain subjects for future investigation and verification.