Wormholes and Averaging over N
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: High Energy Physics: Theory. The day in high-energy theory: gravity, strings, fields and the questions left open.
Kai: We're looking at the paper: "Wormholes and Averaging over N".
Mira: Here's the gist of the paper. Wormholes and Averaging over N introduces "Mellin averaging" as a procedure to define an asymptotic average of functions of N that may reproduce the apparent randomness seen in wormhole physics, provided certain analytic continuation and fluctuation assumptions hold The paper explores how averaging over the integer parameter N might account for wormhole contributions in AdS/CFT, especially when observables fluctuate wildly on superpolynomially small scales in N.
Kai: The Central Puzzle and Proposed Solution Wormholes are contributions to the gravitational path integral that involve a geometrical connection between widely separated regions of spacetime or between two otherwise disconnected spacetimes This introduces randomness, which conflicts with expectations from other quantum gravity theories.
Mira: The authors argue that averaging over N may suffice to account for this randomness if the observable F(N) is defined as a function of a continuous variable N and oscillates wildly on scales vanishing faster than any power of 1/N.
Kai: This requires F(N) to be "wildly oscillatory with oscillations on a scale that vanishes for large N faster than any power of 1/N".
Paper summary: Kai: So, we're looking at this paper called "Wormholes and Averaging over N". The main thing they're tackling is that wormholes seem to introduce randomness into the gravitational path integral. This conflicts with what we usually expect from a definite quantum gravity theory.
Mira: Exactly, and the authors suggest that averaging over the integer parameter N might be enough to reproduce this randomness. They introduce a procedure called "Mellin averaging" to define an asymptotic average of functions that depend on N.
Kai: It's about defining what an average means when you have something that fluctuates wildly on very small scales in N. They claim this works if the dual theory has an analytic continuation in N and those observables fluctuate faster than any power of one over n > one/n <ref:2605.15180#pg1,an analytic continuation in N and>.
Lev: From my side, I gotta ask, is this averaging procedure actually feasible for real hardware? We're talking about superpolynomial times to get the wormholes to dominate, right?
Kai: Right, that's where the paper points—the gains come from being able to assign meaning to those higher order terms in a one over n expansion. It lets you pull out transseries data using the Mellin-Laplace transform of that function.
Mira: And what's interesting is how they use free probability theory to show that joint moments factorize in a way that suggests statistical independence under this specific averaging scheme. It’s a way to mathematically prove the randomness they see in wormhole physics.
Lev: But we need to be careful, because the paper notes that for a simple function like cosine of n times some number, say alpha, the Mellin average just comes out to zero. That suggests it acts like a sequence of independent draws from some distribution on minus one to one.
Conclusion: Kai: So, looking at this work by Kudler-Flam and Witten, the big idea is that we can use Mellin averaging to handle those weird fluctuations caused by wormholes in AdS/CFT. It’s about giving structure to what looks like pure randomness in gravity calculations.
Mira: What they really argue is that for observables fluctuating wildly on superpolynomially small scales, averaging over N isn't just a simple average; it defines a specific type of asymptotic average that mimics the wormhole physics we see.
Kai: It implies that maybe the randomness isn't from some unknown fundamental uncertainty in gravity, but from how we are observing or parameterizing the theory, specifically through this integer N parameter.
Lev: From an engineering standpoint, if this holds up in toy models like chaotic spin chains, it suggests we might be able to simplify those complex path integrals by treating N as a continuous variable with this specific averaging technique instead of dealing with the full complexity at once.
Mira: And they've shown that for general functions, this procedure leads to an N dependent family of probability distributions that eventually settle down into a Gaussian distribution in the large N limit. That convergence is key.
Kai: So, in short, these authors are suggesting we might be able to use this mathematical tool to understand the statistical properties of observables in AdS/CFT more rigorously than just looking at the raw, fluctuating data.
School of Natural Sciences, Institute for Advanced Study, Princeton, New Jersey, USA
hep-th
Submitted: 2026-05-14
Updated: 2026-11-10
Comments: 50 pages, 6 figures; v2: minor errors and typos corrected
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 76/100
The gist: Wormholes and Averaging over N introduces "Mellin averaging" as a procedure to define an asymptotic average of functions of N that may reproduce the apparent randomness seen in wormhole physics…
Key concepts
- Mellin Averaging
- A procedure using the Mellin transform to define an asymptotic average of a function F(N) over integer values of N. It is used to interpret discrete values of N as a sequence of random variables, aiming to model wormhole randomness.
- Spectral Form Factor
- This observable, |ZN (β + iT)|², serves as a prototype for wormhole effects. The analysis suggests that averaging over fluctuations in this factor is necessary to account for the non-trivial 1/N expansion contributions from geometries like the double cone wormhole.
- N-dependent Family of Distributions
- The study finds that F(N) behaves not like draws from one fixed distribution, but rather an N-dependent family of probability distributions, νN. In the large N limit, this family converges to a Gaussian distribution with zero mean and unit variance.
- Free Probability Theory
- This mathematical framework is used to rigorously prove that joint moments factorize under Mellin averaging. This factorization corresponds precisely to the statistical independence observed when applying the averaging procedure.
Terminology
Summary
Wormholes and Averaging over N introduces Mellin averaging
as a procedure to define an asymptotic average of functions of N that may reproduce the apparent randomness seen in wormhole physics, provided certain analytic continuation and fluctuation assumptions hold The paper explores how averaging over the integer parameter N might account for wormhole contributions in AdS/CFT, especially when observables fluctuate wildly on superpolynomially small scales in N.
The Central Puzzle and Proposed Solution
Wormholes are contributions to the gravitational path integral that involve a geometrical connection between widely separated regions of spacetime or between two otherwise disconnected spacetimes This introduces randomness, which conflicts with expectations from other quantum gravity theories. The authors argue that averaging over N may suffice to account for this randomness if the observable F(N) is defined as a function of a continuous variable N and oscillates wildly on scales vanishing faster than any power of 1/N. p. 2 This requires F(N) to be wildly oscillatory with oscillations on a scale that vanishes for large N faster than any power of 1/N
.
Mellin Averaging Procedure
The authors introduce Mellin averaging
as a procedure using the Mellin transform to interpret F(N) at integer values of N as a sequence of random variables. The Mellin average is defined by the formula ⟨F(N)⟩ = X∞ k=0 fk Nk. p. 9 This procedure is motivated by the idea that it defines an asymptotic average rather than a conventional one. p. 2 The Mellin transform of a function F(N) is defined by MF (s) = Z ∞ 1 dN Ns−1F(N). p. 6
Behavior of the Spectral Form Factor
The spectral form factor, ZN (β + iT)2, is used as a prototype for an observable strongly affected by wormholes. The gravitational path integral is interpreted as computing some sort of average, where ⟨ZN (β + iT)2⟩ ≫ ⟨ZN (β + iT)⟩2, suggesting averaging over fluctuations. p. 3 The double cone wormhole contributes a non-trivial 1/N expansion to ZN (β + iT)2. p. 5
Toy Models and Random Matrix Theory
The paper uses toy models, such as the qubit model that can be analytically continued in N, and an explicit construction of a deterministic function of N simulating independent draws from a Gaussian ensemble. In the context of random matrix theory, N is treated as a continuous variable to explore hypotheses about ZN (β + iT). p. 4 Combinatorial approaches using random matrices demonstrate that the connected correlator of resolvents relates to annular noncrossing partitions. p. 1
Conclusion on Statistical Independence
The analysis shows that for a function F(N) = cos(aN) with transcendental a, the Mellin averaging procedure yields ⟨F(N)⟩ = 0, simulating independent draws from the distribution µ on [-1, 1]. This suggests that functions like cos(aN) behave as a succession of independent random draws under this averaging scheme. p. 9 Furthermore, the analysis shows that for general functions F(N), the Mellin average can define an N-dependent family of probability distributions νN that converge to a Gaussian distribution in the large N limit. p. 5 The study concludes that superpolynomial times are needed for Mellin averaging to successfully account for wormholes. p. 2
Where the Gains Come From
The gains come from the ability of Mellin averaging to assign meaning to higher order terms in a 1/N expansion, unlike classical averaging. The procedure allows for the extraction of transseries data from the Mellin-Laplace transform of F(N). p. 9 This method is powerful because it can distinguish between a fixed distribution and an N-dependent family of distributions, which is crucial for understanding the statistical properties of observables in AdS/CFT. The use of free probability theory provides a framework to rigorously prove that joint moments factorize in a way corresponding to statistical independence under Mellin averaging. p. 4
Pruning redundant reasoning tokens
The authors prove the transformation property of the R-transform under free compression by relating it to the conditional expectation E(P(y − PHP)−1P) = (ω(y) − H)−1. This completes the derivation of (4.18), which is key to connecting the spectral form factor and wormhole physics. p. 0 The combinatorial approach provides a more manifest connection to the double cone wormhole geometry. p. 1 The analysis of moments shows that F(N) behaves not as independent draws from a fixed probability distribution ν, but as independent draws from an N-dependent family of probability distributions νN that converge in the large N limit to a Gaussian with zero mean and unit variance. p. 5 The final conclusion is that the double cone geometry can only dominate the gravitational path integral at superpolynomial times. p. 0 The paper also suggests a unique N-dependent reparametrization GN (x) = x+d1(x)N + d2(x)N2 +· · · that maps the N-dependent distribution νN to a fixed Gaussian ensemble. p. 5 The paper also describes toy models of analytic continuation in N using chaotic spin chains. p. 0 The combinatorial approach provides a complementary proof of (4. p. 134) using combinatorial tools. p. 1
How the training signal is built
The training signal is built by defining the Mellin transform M(s) and then performing a contour deformation to define the Mellin average ⟨F(N)⟩ = 1/2πi Z C dsN −sMF (s). The procedure involves exchanging integration over N with a sum over k, leading to an entire function of s for functions like cos(aN). p. 9 The Mellin transform M(s) is shown to be an entire function of s under the condition that Re s < -1/2. p. 8
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Mellin averaging procedure implementation: The AI system can implement a procedure that defines an asymptotic average of a function of N using Mellin transforms, specifically by defining Mellin averaging as
a procedure to define a sort of asymptotic average.
-
Analytic continuation handling: The system can handle analytic continuation in N by treating F(N) as a function of a continuous variable N and exploiting the property that it is
wildly oscillatory with oscillations on a scale that vanishes for large N faster than any power of 1/N.
-
Correlation analysis using spectral form factor: The AI system can analyze correlations between partition functions ZN (beta + iT) for different values of N or T, specifically by checking the criterion that
superpolynomial times are needed for Mellin averaging to successfully account for wormholes.
-
Random Matrix Theory integration: The system can use random matrix theory to explore the behavior of observables like ZN (beta + iT) as a function of N and T, treating N as a continuous variable in this framework.
-
Toy model simulation: The AI system can construct and analyze toy models of deterministic functions F(N) whose values for integer N behave, under Mellin averaging, as Gaussian random variables with non-Gaussian corrections described by an asymptotic series in powers of 1/N.
-
Transseries extraction: The system can apply a Mellin-Laplace transform to the function F(N) to extract transseries data of the form X alpha e - N I alpha X k=0 f (alpha) k Nk, which is used to map erratic functions of N to formal transseries.
-
Spectral form factor behavior prediction: The system can predict that for certain observables like Z∆,N (T)2 at superpolynomial times, the behavior is dominated by a
superpolynomial floor
suggesting thatMellin averaging is potentially a viable resolution to the factorization puzzle.
Sources
- A semiclassical ramp in SYK and in gravity
- The entropy of bulk quantum fields and the entanglement wedge of an evaporating black hole
- Entanglement Wedge Reconstruction and the Information Paradox
- Replica Wormholes and the Entropy of Hawking Radiation
- Replica wormholes and the black hole interior
- N=6 superconformal Chern-Simons-matter theories, M2-branes and their gravity duals
- No Ensemble Averaging Below the Black Hole Threshold
- Towards a holographic description of closed universes
- Emergent Mixed States for Baby Universes and Black Holes
- "Filtering" CFTs at large N: Euclidean Wormholes, Closed Universes, and Black Hole Interiors
- Analytic Continuation Of Chern-Simons Theory
- Comments on the double cone wormhole
- Black Holes and Random Matrices
- Black holes and the butterfly effect
- A precision test of averaging in AdS/CFT
- The double cone geometry is stable to brane nucleation
- Time-periodic solutions in Einstein AdS - massless scalar field system
- All holographic systems have scar states
Related papers
- The facts on quantum self-dual Einstein gravity and self-dual Yang-Mills theory
- Quantum Langlands correspondence for Painlev'e II
- Fortuitous BMN Operators from Multiple Cores
- Building far-from-equilibrium effective field theories using shift symmetries
- Inflation, Cosmic Perturbations and Non-Gaussianities
- Wavefunctions of AdS 3 Universes and T-deformed Torus Partition Functions