OctoInformBeta

OpenAI preprint, September 2026 · §4. The critical block tree · §4

§4.3 (setting): the marked branching law: Doubling block weights and defining ordered trees of connected cores for Fortuin-Kasteleyn parameters p∈(1/2,1)p \in (1/2, 1)

Critical even offspring laws and conditional core mark distributions.

Hover over a blue or green term in the text, or a node in the graph, to find it in the other. Select it to read its declaration both in Lean and in words.

The Setting

In the Fortuin-Kasteleyn (FK) random planar map model with Potts parameter q>4q > 4, set t=qt = \sqrt{q} and p=t/(t+2)∈(1/2,1)p = t/(t+2) \in (1/2, 1). Decomposing a planar map into its blocks yields a plane tree whose vertices have even corner degrees and carry core marks conditioned on their arity. Which critical even offspring distribution governs the corner degrees of this block tree, and how are ordered core trees and arity-conditioned core marks defined?

Offspring Laws and the Inventory Stream

A sequence of block offspring probability weights ν=(νm)m≥0\nu = (\nu_m)_{m \ge 0} is a sequence of nonnegative real numbers satisfying ∑m≥0νm=1\sum_{m \ge 0} \nu_m = 1.✦ The doubled block mass sequence νˆ=(νˆk)k≥0\widehat{\nu} = (\widehat{\nu}_k)_{k \ge 0} of a real sequence ν=(νm)m≥0\nu = (\nu_m)_{m \ge 0} is the real sequence defined by νˆk=νk/2\widehat{\nu}_k = \nu_{k/2} when kk is even and νˆk=0\widehat{\nu}_k = 0 when kk is odd.✦ A critical even offspring distribution π=(πk)k≥0\pi = (\pi_k)_{k \ge 0} is a sequence of nonnegative real numbers with ∑k≥0πk=1\sum_{k \ge 0} \pi_k = 1 that vanishes on odd integers, satisfies π0>0\pi_0 > 0 and π2>0\pi_2 > 0, has mean ∑k≥0kπk=1\sum_{k \ge 0} k \pi_k = 1, and has finite second moment ∑k≥0k2πk<∞\sum_{k \ge 0} k^2 \pi_k < \infty.✦

An FK model parameter tuple PP is a parameter specification consisting of a probability parameter p∈(1/2,1)p \in (1/2, 1), the inventory radius parameter u=(1−p)/16u = (1-p)/16, and the flexible order parameter λP=1+2p/(1−p)\lambda_P = 1 + 2p/(1-p). An inventory order is either a fixed order of one of two burger types or a flexible order✦, and a burger-order alphabet letter is either a burger token of one of those two types or an inventory order token.✦

An inventory stack evaluation state ee is a pair consisting of a list of pending orders and a stack of burger types stored freshest-first.✦ The evaluation of a word ww of alphabet letters on an evaluation state ee is the state obtained by processing the letters of ww from left to right via the single-step transition, in which an order consumes the freshest eligible burger on the stack, returning ee unchanged on the empty word.✦ The word reduction of a word ww of alphabet letters is the evaluation state obtained by evaluating ww starting from empty pending orders and an empty burger stack✦, and a word ww satisfies the empty-reducing word condition if the word reduction of ww leaves both pending orders and the burger stack empty.✦

Under FK model parameters PP with parameter pp, the letter mass function is the function on the five-letter alphabet that assigns mass 1/41/4 to each burger type, p/2p/2 to the flexible order, and (1−p)/4(1-p)/4 to each fixed order type✦, and the letter distribution is the probability distribution on the five-letter alphabet with these masses.✦ For any index set II, the independent and identically distributed (i.i.d.) inventory stream measure is the product probability measure on sequences (xi)i∈I(x_i)_{i \in I} of independent copies of the letter distribution; taking I=ℕI = \mathbb{N} gives the measure ℙP\mathbb{P}_P on streams x=(xi)i≥0x = (x_i)_{i \ge 0}. For each N≥0N \ge 0, the letter stream initial word wN(x)w_N(x) of a stream xx is the prefix (x0,…,xN−1)(x_0, \dots, x_{N-1}) of length NN✦, and the empty-reducing word probability ηN(P)\eta_N(P) is the probability given by (1): ηN(P)=ℙP(wN(x) satisfies the empty-reducing word condition),\eta_N(P) = \mathbb{P}_P\bigl(w_N(x) \text{ satisfies the empty-reducing word condition}\bigr),(1)✦ and the total empty word probability τ(P)\tau(P) is the series defined by (2): τ(P)=∑n≥0η2n(P).\tau(P) = \sum_{n \ge 0} \eta_{2n}(P).(2)✦

Typed Cores and Core Labels

A typed burger-order letter is either a burger token or an order token, each carrying one of the two burger types.✦ For any finite alphabet 𝒜\mathcal{A} and integer N≥0N \ge 0, the set of words of length NN over 𝒜\mathcal{A} is the finite set of all NN-letter sequences in 𝒜\mathcal{A}; write 𝒲N\mathcal{W}_N when 𝒜\mathcal{A} is the typed alphabet.✦ Typed word stack execution is the operation that processes a typed word against a freshest-first stack of burger types, returning either the resulting stack or failure upon an illegal closure✦; for each integer N≥0N \ge 0, the empty-reducing typed word set ℰN\mathcal{E}_N is the subset of 𝒲N\mathcal{W}_N consisting of the words of length NN whose execution from an empty stack returns an empty stack.✦ Given a typed word ww and an initial stack ss of burger types, the flexible closing count f(w,s)f(w, s) is the number of flexible order closures encountered when processing ww against ss✦; writing f0(w)f_0(w) when ss is the empty stack, the inventory word monomial weight of ww at parameters λ\lambda and rr is λf0(w)r|w|\lambda^{f_0(w)} r^{|w|}, where |w||w| is the length of ww.✦

A non-crossing arch matching EE on a list ll of positions is an arch list of ordered pairs formed inductively from the empty matching on the empty list by either skipping an unmatched element of ll or pairing two elements (i,j)(i, j) around non-crossing arch matchings on the interior segment between ii and jj and the exterior segment after jj, with (i,j)(i, j) preceding the combined interior and exterior arches.✦ For a non-crossing arch matching EE, the endpoints of EE are the list obtained by flattening every arch (i,j)∈E(i, j) \in E into its two coordinates i,ji, j✦, and the arch transposition product of EE is the permutation defined recursively as the identity on the empty list and, for a leading arch (i,j)(i, j) followed by tail E′E', as the product of the arch transposition product of E′E' with the transposition (ij)(i\; j). Given a typed-letter assignment vv on positions and a position list ll, a two-stack matching certificate MM is a pair (E1,E2)(E_1, E_2) of non-crossing arch matchings on ll whose concatenated endpoints form a permutation of ll, where each arch in E1E_1 pairs a burger with an order of the first type and each arch in E2E_2 pairs a burger with an order of the second type✦; the stack matching mate permutation σM\sigma_M of MM is the arch transposition product of the concatenation of E1E_1 and E2E_2.✦

For a permutation ρ\rho on {0,…,N−1}\{0, \dots, N-1\}, a chord opening endpoint (opener) ii of ρ\rho is an index in {0,…,N−1}\{0, \dots, N-1\} satisfying i<ρ(i)i < \rho(i). Two openers xx and yy of ρ\rho satisfy the chord crossing relation if either x<y<ρ(x)<ρ(y)x < y < \rho(x) < \rho(y) or y<x<ρ(y)<ρ(x)y < x < \rho(y) < \rho(x)✦, and the chord intersection graph GρG_\rho of ρ\rho is the simple graph on the openers of ρ\rho whose edges are the crossing pairs.✦ A typed word ww satisfies the connected core word condition if there exist K≥0K \ge 0, a typed-letter assignment vv on {0,…,K−1}\{0, \dots, K-1\} whose values along (0,…,K−1)(0, \dots, K-1) form ww, and a two-stack matching certificate MM for vv along (0,…,K−1)(0, \dots, K-1) whose mate permutation σM\sigma_M has 00 as an opener from which every opener of σM\sigma_M is reachable in GσMG_{\sigma_M}. The set 𝒞\mathcal{C} of connected typed cores is the set of typed words satisfying the connected core word condition.✦

Given a real parameter λ\lambda and an integer m≥0m \ge 0, the core generating polynomial coefficient cm(λ)c_m(\lambda) is the real sum defined by (3): cm(λ)=∑w∈ℰ2m∩𝒞λf0(w),c_m(\lambda) = \sum_{w \in \mathcal{E}_{2m} \cap \mathcal{C}} \lambda^{f_0(w)},(3)✦ and the block generating polynomial coefficient bm(λ)b_m(\lambda) is the real coefficient given by (4): bm(λ)={1+c0(λ),m=0,cm(λ),m≥1.b_m(\lambda) = \begin{cases} 1 + c_0(\lambda), & m = 0, \\ c_m(\lambda), & m \ge 1. \end{cases}(4)✦ For FK model parameters PP and size m≥0m \ge 0, the actual FK block mass νP(m)\nu_P(m) is the real number defined by (5): νP(m)=bm(λP)(uτ(P)2)mτ(P).\nu_P(m) = \frac{b_m(\lambda_P)\,\bigl(u\,\tau(P)^2\bigr)^m}{\tau(P)}.(5)✦

An optional core label ℓ\ell is an element of 𝒞∪{⌀}\mathcal{C} \cup \{\varnothing\}, representing either a connected typed core c∈𝒞c \in \mathcal{C} or the empty label ⌀\varnothing.✦ The arity arity⁡(ℓ)\operatorname{arity}(\ell) of an optional core label ℓ\ell is the integer defined by arity⁡(⌀)=0\operatorname{arity}(\varnothing) = 0 and arity⁡(c)=|c|\operatorname{arity}(c) = |c| for c∈𝒞c \in \mathcal{C}. For real parameters λ\lambda, rr, and τ\tau and an optional core label ℓ\ell, the parametric core label mass ω(λ,r,τ;ℓ)\omega(\lambda, r, \tau; \ell) is the real number defined by (6): ω(λ,r,τ;ℓ)={1/τ,ℓ=⌀,λf0(c)(rτ)|c|/τ,ℓ=c∈𝒞.\omega(\lambda, r, \tau; \ell) = \begin{cases} 1/\tau, & \ell = \varnothing, \\ \lambda^{f_0(c)}\,(r\tau)^{|c|}/\tau, & \ell = c \in \mathcal{C}. \end{cases}(6)✦ Under FK model parameters PP, the root core mark mass ωP(ℓ)\omega_P(\ell) of an optional core label ℓ\ell is ω(λP,u,τ(P);ℓ)\omega\bigl(\lambda_P, \sqrt{u}, \tau(P);\, \ell\bigr)✦, and the root core mark distribution ΛP\Lambda_P is the probability distribution on 𝒞∪{⌀}\mathcal{C} \cup \{\varnothing\} assigning probability mass max⁡(ωP(ℓ),0)\max(\omega_P(\ell), 0) to each label ℓ\ell.✦

The Definition

Definition 1. Let ν=(νm)m≥0\nu = (\nu_m)_{m \ge 0} be a sequence of block offspring probability weights satisfying ν0>0\nu_0 > 0, ν1>0\nu_1 > 0, ∑m≥0mνm=1/2\sum_{m \ge 0} m \nu_m = 1/2, and ∑m≥0m2νm<∞\sum_{m \ge 0} m^2 \nu_m < \infty. The constructed even offspring distribution π(ν)=(πk(ν))k≥0\pi(\nu) = (\pi_k(\nu))_{k \ge 0} associated with ν\nu is the critical even offspring distribution given by (7): πk(ν)=νˆk(k≥0).\pi_k(\nu) = \widehat{\nu}_k \qquad (k \ge 0).(7)✦

Under FK model parameters PP with p∈(1/2,1)p \in (1/2, 1), the critical FK corner offspring law πP\pi_P is the critical even offspring distribution (8) obtained by applying the construction (7) to the actual FK block mass νP\nu_P: πP(k)=πk(νP)(k≥0).\pi_P(k) = \pi_k(\nu_P) \qquad (k \ge 0).(8)✦

The family 𝒯core\mathcal{T}_{\mathrm{core}} of ordered trees of cores is the inductively generated set defined by (9), where ∘\circ denotes a leaf node and each element in the component indexed by c∈𝒞c \in \mathcal{C} is a branch node carrying core cc and an ordered |c||c|-tuple of subtrees indexed by the corners of cc: 𝒯core={∘}⊔⨆c∈𝒞({c}×𝒯core|c|).\mathcal{T}_{\mathrm{core}} = \{\circ\} \sqcup \bigsqcup_{c \in \mathcal{C}} \bigl(\{c\} \times \mathcal{T}_{\mathrm{core}}^{|c|}\bigr).(9)✦

Writing δ⌀\delta_{\varnothing} for the point mass at the empty label ⌀\varnothing, the conditional core mark distribution ΛP,k\Lambda_{P, k} on 𝒞∪{⌀}\mathcal{C} \cup \{\varnothing\} for FK model parameters PP and arity k≥0k \ge 0 is the probability distribution defined by (10): ΛP,k(ℓ)={ΛP(ℓ|arity⁡(ℓ)=k),if there exists ℓ′∈𝒞∪{⌀} with arity⁡(ℓ′)=k,δ⌀(ℓ),otherwise.\Lambda_{P, k}(\ell) = \begin{cases} \Lambda_P\bigl(\ell \;\big|\; \operatorname{arity}(\ell) = k\bigr), & \text{if there exists } \ell' \in \mathcal{C} \cup \{\varnothing\} \text{ with } \operatorname{arity}(\ell') = k, \\ \delta_{\varnothing}(\ell), & \text{otherwise}. \end{cases}(10)✦

What Is Checked

The offspring distributions (7) and (8) are stated over block offspring probability weights ν\nu satisfying ν0>0\nu_0 > 0, ν1>0\nu_1 > 0, ∑m≥0mνm=1/2\sum_{m \ge 0} m \nu_m = 1/2, and ∑m≥0m2νm<∞\sum_{m \ge 0} m^2 \nu_m < \infty, as well as FK model parameters PP with p∈(1/2,1)p \in (1/2, 1) and the actual FK block mass νP\nu_P built from the inventory stream and core generating polynomial coefficients. The ordered tree of cores (9) is stated over connected typed cores c∈𝒞c \in \mathcal{C} defined by two-stack matching certificates and chord intersection graphs, while the conditional core mark distribution (10) is stated over arities k≥0k \ge 0 and the root core mark distribution ΛP\Lambda_P (for FK model parameters PP) on optional core labels. For any FK model parameters PP, the actual FK block mass satisfies νP(m)≥0\nu_P(m) \ge 0 for every m≥0m \ge 0✦, with νP(0)>0\nu_P(0) > 0✦ and νP(1)>0\nu_P(1) > 0✦.