Adelic Models of Percolation

✨

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: Connecting Two Different Worlds

Imagine you are trying to understand how a rumor spreads through a crowd. You have two very different ways of modeling this crowd:

The Grid City (Ordinary Lattices): Imagine a city laid out in perfect square blocks (like Manhattan). People live at the intersections. A rumor can jump from one person to another, but it's more likely to jump to a neighbor nearby than to someone across town. However, there is a tiny chance it can jump anywhere. This is the "Ordinary Lattice" model.
The Russian Nesting Doll (Hierarchical Lattices): Now imagine a crowd organized like a set of Russian nesting dolls. Everyone belongs to a small group, those groups belong to bigger groups, and so on. The "distance" between two people isn't measured in miles, but by how deep you have to go into the nesting dolls to find a common ancestor. This is the "Hierarchical Lattice" model.

The Problem: Mathematicians have studied both models separately for decades. They know a lot about the Grid City, but the Russian Dolls are easier to calculate with. The big question is: Can we use the easy Russian Doll model to predict what happens in the complex Grid City?

The Solution: This paper says "Yes," but to do it, we need a secret translator. That translator is a concept from advanced number theory called "Adelic Geometry."

The Secret Translator: The "Adelic Shadow"

The paper uses a brilliant idea proposed by the late mathematician Yuri Manin. He suggested that for every physical system we understand using real numbers (like our Grid City), there is a "shadow" version made of numbers that behave differently (like prime numbers and their powers).

Think of it like this:

The Real World: You have a smooth, continuous river (the Grid City).
The Shadow World: You have a river made of distinct, separate pebbles (the Hierarchical Dolls).

The paper builds a bridge between these two rivers using a mathematical tool called the Adelic Product Formula. This formula is like a universal exchange rate that says: "If you multiply all the values of a number across every possible 'shadow' world, you get the inverse of its value in the real world."

By using this exchange rate, the author shows that the behavior of the rumor in the Russian Dolls is mathematically linked to the behavior of the rumor in the Grid City.

The Three Steps of the Bridge

The paper constructs a three-step bridge to connect the two worlds.

Step 1: The "Power Mean" Deformation (The Stretchy Rubber Band)

First, the author creates a "stretchy" version of the Grid City.

Analogy: Imagine the Grid City is a rubber sheet. Usually, the chance of a rumor jumping depends on the straight-line distance (Euclidean distance).
The Trick: The author introduces a "knob" (a parameter called $t$ $t$ ).
- Turn the knob to 2, and you get the standard Grid City (straight lines).
- Turn the knob to 0, and the distance changes to something called "Toric Volume." This is like measuring distance by the product of coordinates rather than the sum. It's a weird, multi-dimensional way of measuring space that sounds like a "Toric Percolation" model.
Why? This "Toric" model is the perfect middle ground. It's still a Grid City, but it looks mathematically similar to the Russian Dolls.

Step 2: The Function Field Bridge (The Finite World)

Next, the author looks at the Russian Dolls (Hierarchical Lattices) and realizes they are actually just a specific view of a "Function Field" (a type of number system based on polynomials, like $x^2 + 1$ ).

Analogy: Think of the Russian Dolls as a giant tree. The author shows that this tree is actually the "top" of a mathematical structure built from polynomials.
The Magic: Using the Adelic formula, the author proves that the "top" of this tree (the Hierarchical Lattice) is mathematically equivalent to the "bottom" of the tree (the finite parts of the polynomial system).
Result: We can now translate the rules of the Russian Dolls into the language of polynomials.

Step 3: The Number Field Bridge (The Integer World)

Finally, the author connects the polynomial world (Step 2) to the Grid City (Step 1) using Number Fields (systems based on integers, like the ring of integers in a complex number system).

Analogy: Imagine the Grid City is built on a lattice of integers. The author shows that if you look at this lattice through the lens of "Adelic Geometry," the "finite" parts (the non-Archimedean places) look exactly like the polynomial world we built in Step 2.
The Connection:
1. The Hierarchical Lattice (Russian Dolls) is the "top" of the Polynomial world.
2. The Toric Percolation (the stretchy Grid City) is the "top" of the Integer world.
3. The "bottoms" of both worlds (the finite parts) are identical.

Because the "bottoms" are the same, the "tops" must be related.

Why Does This Matter? (The "So What?")

In the world of physics and math, the Grid City is the realistic model of how things like electricity, magnetism, or diseases spread in our 3D world. However, it is incredibly hard to solve the math for the Grid City.

The Russian Dolls are a simplified, artificial model. They are easy to solve, but they don't look like our real world.

The Breakthrough:
This paper proves that you can take the easy answers you get from the Russian Dolls, run them through this "Adelic Bridge," and get accurate predictions for the complex Grid City.

It's like saying: "I can't calculate the weather for the whole Earth (Grid City), but I can calculate the weather for a perfect, simplified sphere (Russian Dolls). Because of this new bridge, I can now use my sphere calculation to predict the Earth's weather with surprising accuracy."

Summary in a Nutshell

The Goal: Connect a realistic, messy model of spreading (Grid) with a simplified, fractal model (Dolls).
The Tool: Use "Adelic Geometry," a way of looking at numbers that treats "real" numbers and "prime" numbers as two sides of the same coin.
The Method:
- Stretch the Grid City until it looks like the Dolls (using Power Means).
- Show that the Dolls are just the "top" of a polynomial system.
- Show that the Stretched Grid is the "top" of an integer system.
- Prove the "bottoms" of the polynomial and integer systems are twins.
The Result: We can now use the easy math of the fractal world to solve the hard math of the real world.

This work honors Yuri Manin's vision that deep number theory (the study of prime numbers and integers) holds the keys to understanding the fundamental laws of physics.

1. Problem Statement

The paper addresses the geometric and probabilistic relationship between two distinct classes of long-range percolation models in statistical physics:

Ordinary Lattice Percolation: Models defined on lattices $\Lambda \subset \mathbb{R}^d$ (specifically rings of integers $\mathcal{O}_K$ of number fields embedded via the Minkowski embedding). The connection probability between vertices $x, y$ decays as a power of the Euclidean distance: $1 - \exp(-\beta \|x-y\|^{-d-\alpha})$ .
Hierarchical Lattice Percolation: Models defined on abelian groups with a self-similar, fractal-like ultrametric geometry (e.g., $H^d_L = \oplus_{i=1}^\infty (\mathbb{Z}/L\mathbb{Z})^d$ ).

While recent work (e.g., by Hutchcroft) has shown that hierarchical models can sometimes bound or estimate critical exponents for ordinary lattices, a direct geometric unification of these two systems has been lacking. The paper seeks to establish a rigorous "adelic" bridge between them, realizing both as different "fibers at infinity" of a unified arithmetic construction based on global fields (number fields and function fields).

2. Methodology

The author employs a strategy inspired by Yuri Manin's "adelic product formula," which relates Archimedean (real/complex) and non-Archimedean ( $p$ -adic) norms. The methodology involves three main steps:

A. Deformation via Power Mean Functions

To connect the standard Euclidean lattice model to an intermediate model, the author introduces a 1-parameter family of percolation models $PM_t(\Lambda)$ on a lattice $\Lambda$ .

The connection probability is governed by a function $J_{\alpha, t, \lambda}(x-y) = M_t(\lambda, x-y)^{-N-\alpha}$ , where $M_t$ is the power mean function.
Key Special Cases:
- $t=2$ : Recovers the standard Euclidean long-range percolation model ( $J_\alpha \sim \|x-y\|^{-N-\alpha}$ ).
- $t=0$ : Recovers the Toric Volume Percolation model, where the connection probability is governed by the product of coordinates (related to the Haar measure on the algebraic torus).
Monotonicity of the power mean function allows for inequalities relating the critical temperatures of these models.

B. Arithmetic Geometry and Global Fields

The paper constructs adelic models over two types of global fields:

Function Fields ( $K = \mathbb{F}_q(C)$ ):
- The infinite place ( $\infty$ ) corresponds to the hierarchical lattice model. The polynomial ring $\mathbb{F}_q[t]$ is isomorphic to the hierarchical lattice $H^1_q$ .
- The finite places correspond to local models where the connection probability is non-integrable, guaranteeing an infinite cluster for any $\beta > 0$ .
- The Adelic Product Formula ( $\prod |f|_x = 1$ ) relates the behavior at the infinite place (hierarchical) to the product of behaviors at finite places.
Number Fields ( $K$ ):
- The Archimedean places (via Minkowski embedding) correspond to the Toric Percolation model on the ring of integers $\mathcal{O}_K$ .
- The non-Archimedean places (finite places) correspond to local percolation models on subsets of $\mathcal{O}_K$ defined by terminating or periodic $\mathfrak{p}$ -adic expansions. These local models are geometrically described using Bruhat-Tits trees and also exhibit non-integrability (infinite clusters for all $\beta > 0$ ).
- The Adelic Product Formula for number fields relates the Archimedean (Toric) behavior to the product of non-Archimedean behaviors.

C. The Adelic Diagram

The core of the methodology is a commutative diagram (Equation 1.2) linking:

Hierarchical Lattice $\leftrightarrow$ Function Field Infinite Place.
Function Field Finite Places $\leftrightarrow$ Number Field Non-Archimedean Places (via local Bruhat-Tits tree geometry).
Number Field Archimedean Places $\leftrightarrow$ Toric Percolation $\leftrightarrow$ Standard Lattice Percolation (via the power mean deformation $t=0 \to t=2$ ).

3. Key Contributions

Unification of Geometries: The paper provides the first explicit arithmetic framework showing that long-range percolation on ordinary lattices and hierarchical lattices are "adelic shadows" of the same underlying global field structures.
Toric Percolation Model: Introduction of a new specific percolation model based on the toric volume form (product of coordinates), which serves as the crucial intermediate step ( $t=0$ ) between the hierarchical model and the standard Euclidean model ( $t=2$ ).
Bruhat-Tits Tree Interpretation: A geometric reinterpretation of local non-Archimedean percolation models using the boundary of Bruhat-Tits trees, clarifying why these local models always possess infinite clusters (due to non-integrability of the connection kernel).
Critical Temperature Estimates: Derivation of rigorous inequalities relating the critical inverse temperatures ( $\beta_c$ ) of the different models using the Dedekind zeta function $\zeta_K(s)$ and the Riemann zeta function.

4. Key Results

Theorem 4.9 & 5.17 (Adelic Comparison): The paper establishes that the percolation probability of the hierarchical model (function field infinity) and the toric/lattice model (number field infinity) can be bounded by the adelic models over the finite places.
- Specifically, for a number field $K$ , the probability of connection in the adelic model $P_{K, \text{fin}}$ is bounded by the toric model $P_T$ scaled by factors involving the Dedekind zeta function $\zeta_K(\beta)$ .
- The bounds depend on the choice of parameters $\alpha$ and the presence of Siegel zeros (real zeros of $\zeta_K(s)$ near $s=1$ ), which restrict the valid range of $\beta$ for the estimates.
Proposition 4.4 & 5.9 (Local Non-Integrability): It is proven that for local models at finite places (both function and number fields), the connection kernel is non-integrable. Consequently, the critical inverse temperature is $\beta_c = 0$ ; an infinite cluster exists with probability 1 for any $\beta > 0$ .
Proposition 5.18 (Critical Temperature Bounds):
- If $\beta > \max\{1, \beta_{c,T} \cdot \zeta_K(\beta)\}$ , an infinite cluster exists in the adelic model.
- If $\beta \leq \min\{\beta_0, \beta_{c,T} \cdot \zeta_K(\beta)\}$ (where $\beta_0$ avoids Siegel zeros), no infinite cluster exists.
- This provides a method to estimate the critical exponent of the ordinary lattice model using the arithmetic properties of the number field.

5. Significance

Interdisciplinary Synthesis: The work successfully merges Statistical Physics (percolation, critical phenomena) with Arithmetic Geometry (adeles, zeta functions, Bruhat-Tits trees). It validates Manin's conjecture that arithmetic concepts (like the adelic product formula) can provide deep insights into physical models.
New Analytical Tools: By translating percolation problems into the language of global fields, the paper opens the door to using powerful tools from analytic number theory (e.g., properties of $\zeta_K(s)$ , distribution of primes) to solve problems in statistical mechanics, such as estimating critical exponents for long-range percolation.
Geometric Insight: The identification of hierarchical lattices with the "fiber at infinity" of function fields and ordinary lattices with the "fiber at infinity" of number fields offers a profound geometric perspective on why these seemingly different systems share similar scaling behaviors.
Future Directions: The framework suggests that the critical behavior of percolation on lattices associated with specific number fields (e.g., cyclotomic fields) might be directly computable or bounded using the specific arithmetic invariants of those fields.

In summary, Marcolli demonstrates that the "long-range" nature of percolation on both ordinary and hierarchical lattices is a manifestation of the same adelic structure, where the Archimedean (real) and non-Archimedean ( $p$ -adic) components are linked by the product formula, allowing for the transfer of results between these distinct geometric settings.