OctoInformBeta

OpenAI preprint, September 2026 · §7. Pair rings and weighted Hilbert lower bounds

§7 (setting): Quotient rings for root and cone pair relations together with integer-weighted polynomial and real-weighted quotient subspaces

Pair rings and weighted Hilbert functions.

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

Given a commutative ring RR, nonnegative integers dd and kk, and elements q1,…,qd∈Rq_1, \dots, q_d \in R, consider the system of equations aikbi=qia_i^k b_i = q_i (1≤i≤d1 \le i \le d) in variables ai,bia_i, b_i. Algebraic approaches to such systems work inside the coordinate ring of solutions and measure the growth of functions on it by filtering polynomial and quotient rings according to variable weights. These notes set out the corresponding quotient rings together with the weighted spaces and Hilbert functions.

Relations and Weights

For a commutative ring RR, nonnegative integers dd and kk, and a tuple q=(q1,…,qd)∈Rdq = (q_1, \dots, q_d) \in R^d, the ii-th pair ring relation rk,q,ir_{k,q,i} (1≤i≤d1 \le i \le d) is the polynomial in R[a1,…,ad,b1,…,bd]R[a_1, \dots, a_d, b_1, \dots, b_d] given by (1): rk,q,i=aikbi−qi.r_{k,q,i} = a_i^k b_i - q_i.(1)✦

Given a field KK, a nonnegative integer dd, and θ=(θ1,…,θd)∈Kd\theta = (\theta_1, \dots, \theta_d) \in K^d, the root pair polynomials qθ,irootq^{\mathrm{root}}_{\theta,i} (1≤i≤d1 \le i \le d) are the polynomials in K[t]K[t] given by (2), forming the tuple qθroot=(qθ,1root,…,qθ,droot)∈K[t]dq^{\mathrm{root}}_{\theta} = (q^{\mathrm{root}}_{\theta,1}, \dots, q^{\mathrm{root}}_{\theta,d}) \in K[t]^d: qθ,iroot(t)=t−θi.q^{\mathrm{root}}_{\theta,i}(t) = t - \theta_i.(2)✦

When KK is a field, d,k,A,Bd, k, A, B are nonnegative integers, and θ=(θ1,…,θd)∈Kd\theta = (\theta_1, \dots, \theta_d) \in K^d, the ii-th pair cone polynomial qk,A,B,θ,iconeq^{\mathrm{cone}}_{k,A,B,\theta,i} (1≤i≤d1 \le i \le d) is the polynomial in K[t0,t1,t2]K[t_0, t_1, t_2] defined by (3), with tuple qk,A,B,θcone=(qk,A,B,θ,1cone,…,qk,A,B,θ,dcone)∈K[t0,t1,t2]dq^{\mathrm{cone}}_{k,A,B,\theta} = (q^{\mathrm{cone}}_{k,A,B,\theta,1}, \dots, q^{\mathrm{cone}}_{k,A,B,\theta,d}) \in K[t_0, t_1, t_2]^d: qk,A,B,θ,icone(t0,t1,t2)=t0kt1+θit2kA+B.q^{\mathrm{cone}}_{k,A,B,\theta,i}(t_0, t_1, t_2) = t_0^k t_1 + \theta_i t_2^{kA + B}.(3)✦

If σ\sigma is a finite variable set, ω:σ→ℝ\omega : \sigma \to \mathbb{R} is a real weight function, and a=(ai)i∈σ∈ℤ≥0σa = (a_i)_{i \in \sigma} \in \mathbb{Z}_{\ge 0}^{\sigma} is a nonnegative integer exponent vector, the monomial weight wtω(a)\mathrm{wt}_{\omega}(a) is the real number defined by (4): wtω(a)=∑i∈σω(i)ai.\mathrm{wt}_{\omega}(a) = \sum_{i \in \sigma} \omega(i)\, a_i.(4)✦

For a finite variable set σ\sigma, a real weight function ω:σ→ℝ\omega : \sigma \to \mathbb{R}, and a real threshold T∈ℝT \in \mathbb{R}, the weighted exponent bound Bω,TB_{\omega,T} is the exponent vector in ℤ≥0σ\mathbb{Z}_{\ge 0}^{\sigma} defined at each i∈σi \in \sigma by (5): Bω,T(i)=⌊T/ω(i)⌋+,B_{\omega,T}(i) = \lfloor T / \omega(i) \rfloor_+,(5) where ⌊x⌋+=max⁡(0,⌊x⌋)\lfloor x \rfloor_+ = \max(0, \lfloor x \rfloor) denotes the nonnegative integer floor of a real number xx.

Given a finite variable set σ\sigma, real weight function ω:σ→ℝ\omega : \sigma \to \mathbb{R}, and real threshold T∈ℝT \in \mathbb{R}, the weighted monomials ℳω(T)\mathcal{M}_{\omega}(T) form a finite set of nonnegative integer exponent vectors in ℤ≥0σ\mathbb{Z}_{\ge 0}^{\sigma} associated with ω\omega and TT.

The Definition

Definition 1. For a commutative ring RR, nonnegative integers dd and kk, and a tuple q=(q1,…,qd)∈Rdq = (q_1, \dots, q_d) \in R^d, the pair ring 𝒫k,q(R)\mathcal{P}_{k,q}(R) is the quotient ring of R[a1,…,ad,b1,…,bd]R[a_1, \dots, a_d, b_1, \dots, b_d] defined by (6): 𝒫k,q(R)=R[a1,…,ad,b1,…,bd]/⟨rk,q,i:1≤i≤d⟩.\mathcal{P}_{k,q}(R) = R[a_1, \dots, a_d, b_1, \dots, b_d] \big/ \langle r_{k,q,i} : 1 \le i \le d \rangle.(6)✦

For a commutative ring RR, nonnegative integers dd and kk, and q∈Rdq \in R^d, the pair ring coefficient map ιk,q:R→𝒫k,q(R)\iota_{k,q} : R \to \mathcal{P}_{k,q}(R) is the ring homomorphism sending each r∈Rr \in R to the residue class of the constant polynomial rr.✦

When KK is a field, dd and kk are nonnegative integers, and θ∈Kd\theta \in K^d, the root pair ring ℛk,θ(K)\mathcal{R}_{k,\theta}(K) is the pair ring over K[t]K[t] given by (7): ℛk,θ(K)=𝒫k,qθroot(K[t]).\mathcal{R}_{k,\theta}(K) = \mathcal{P}_{k,\, q^{\mathrm{root}}_{\theta}}(K[t]).(7)✦

Take a field KK, nonnegative integers d,k,A,Bd, k, A, B, and θ∈Kd\theta \in K^d; the pair cone ring 𝒞k,A,B,θ(K)\mathcal{C}_{k,A,B,\theta}(K) is the pair ring 𝒫k,qk,A,B,θcone(K[t0,t1,t2])\mathcal{P}_{k,\, q^{\mathrm{cone}}_{k,A,B,\theta}}(K[t_0, t_1, t_2]).✦

Over a field KK, for nonnegative integers d,k,A,Bd, k, A, B and θ∈Kd\theta \in K^d, the affine pair ring 𝒜k,A,B,θ(K)\mathcal{A}_{k,A,B,\theta}(K) is the quotient ring of 𝒞k,A,B,θ(K)\mathcal{C}_{k,A,B,\theta}(K) defined by (8): 𝒜k,A,B,θ(K)=𝒞k,A,B,θ(K)/⟨ιk,qk,A,B,θcone(t2)−1⟩.\mathcal{A}_{k,A,B,\theta}(K) = \mathcal{C}_{k,A,B,\theta}(K) \big/ \langle \iota_{k,\, q^{\mathrm{cone}}_{k,A,B,\theta}}(t_2) - 1 \rangle.(8)✦

Given an index set σ\sigma with variables (xi)i∈σ(x_i)_{i \in \sigma}, a commutative semiring KK, an integer weight function w:σ→ℤ≥0w : \sigma \to \mathbb{Z}_{\ge 0}, and an integer threshold T∈ℤ≥0T \in \mathbb{Z}_{\ge 0}, the space of weighted polynomials Wσ,K(w,T)W_{\sigma,K}(w, T) is the integer-weighted KK-submodule of K[σ]K[\sigma] defined by (9), where m:σ→ℤ≥0m : \sigma \to \mathbb{Z}_{\ge 0} ranges over finitely supported exponent vectors and xm=∏i∈σxim(i)x^m = \prod_{i \in \sigma} x_i^{m(i)}: Wσ,K(w,T)=span⁡K{xm∈K[σ]:∑i∈σw(i)m(i)≤T}.W_{\sigma,K}(w, T) = \operatorname{span}_K \left\{ x^m \in K[\sigma] : \sum_{i \in \sigma} w(i)\, m(i) \le T \right\}.(9)✦

Suppose σ\sigma is a finite variable set with indeterminates (xi)i∈σ(x_i)_{i \in \sigma}, KK is a field, I⊆K[σ]I \subseteq K[\sigma] is an ideal, ω:σ→ℝ\omega : \sigma \to \mathbb{R} is a real weight function, and T∈ℝT \in \mathbb{R}; then the space of weighted functions ℱI,ω(T)\mathcal{F}_{I,\omega}(T) is the real-weighted KK-subspace of K[σ]/IK[\sigma]/I defined by (10), where xa=∏i∈σxiaix^a = \prod_{i \in \sigma} x_i^{a_i}: ℱI,ω(T)=span⁡K{xamod⁡I:a∈ℳω(T)}.\mathcal{F}_{I,\omega}(T) = \operatorname{span}_K \left\{ x^a \bmod I : a \in \mathcal{M}_{\omega}(T) \right\}.(10)✦

For a finite variable set σ\sigma, a field KK, an ideal I⊆K[σ]I \subseteq K[\sigma], a real weight function ω:σ→ℝ\omega : \sigma \to \mathbb{R}, and a real threshold T∈ℝT \in \mathbb{R}, the weighted Hilbert function HI,ω(T)H_{I,\omega}(T) is the KK-dimension defined by (11): HI,ω(T)=dim⁡KℱI,ω(T).H_{I,\omega}(T) = \dim_K \mathcal{F}_{I,\omega}(T).(11)✦

In the paper, the result reads: filtered coordinate rings and weighted Hilbert functions HI,ω(T)H_{I,\omega}(T) for the dd-equation system aikbi=n−θia_i^k b_i = n - \theta_i (1≤i≤d1 \le i \le d), where nn stands for tt, and its homogenized cone with positive weights (U,V)(U, V).

What Is Checked

The definitions are stated over an arbitrary commutative ring RR with d,k∈ℤ≥0d, k \in \mathbb{Z}_{\ge 0}, q∈Rdq \in R^d, and relations rk,q,ir_{k,q,i} (1≤i≤d1 \le i \le d) for 𝒫k,q(R)\mathcal{P}_{k,q}(R); over a field KK with d,k,A,B∈ℤ≥0d, k, A, B \in \mathbb{Z}_{\ge 0}, θ∈Kd\theta \in K^d, qθrootq^{\mathrm{root}}_{\theta}, qk,A,B,θconeq^{\mathrm{cone}}_{k,A,B,\theta}, 𝒞k,A,B,θ(K)\mathcal{C}_{k,A,B,\theta}(K), and ιk,qk,A,B,θcone\iota_{k,\, q^{\mathrm{cone}}_{k,A,B,\theta}} for ℛk,θ(K)\mathcal{R}_{k,\theta}(K) and 𝒜k,A,B,θ(K)\mathcal{A}_{k,A,B,\theta}(K); over an arbitrary index set σ\sigma and commutative semiring KK with w:σ→ℤ≥0w : \sigma \to \mathbb{Z}_{\ge 0} and T∈ℤ≥0T \in \mathbb{Z}_{\ge 0} for Wσ,K(w,T)W_{\sigma,K}(w, T); and over a finite variable set σ\sigma, a field KK, an ideal I⊆K[σ]I \subseteq K[\sigma], ω:σ→ℝ\omega : \sigma \to \mathbb{R}, T∈ℝT \in \mathbb{R}, and ℳω(T)\mathcal{M}_{\omega}(T) for ℱI,ω(T)\mathcal{F}_{I,\omega}(T) and HI,ω(T)H_{I,\omega}(T).