Universal backward-geodesic endpoint argmax conjecture

Consider an exclusion process with generator specified by the paper and initial condition Xn(0)=n/ρX_n(0)=-\lfloor n/\rho\rfloor for some ρ(0,1)\rho\in(0,1). Assume its translation-invariant stationary measure μρ\mu_\rho satisfies the paper's Assumption~. Let J(ρ)J(\rho) be the stationary current and let A(ρ)A(\rho) and Γ(ρ)\Gamma(\rho) be the associated KPZ parameters. Let N(t0)N(t\downarrow0) denote the backward-geodesic endpoint index, and define

u^=argmaxuR{A2(u)u2}.\hat u=\operatorname*{argmax}_{u\in\mathbb{R}}\left\{{\cal A}_2(u)-u^2\right\}.

Universal backward-geodesic endpoint conjecture. There exists a universal constant c2c_2 such that

limtXN(t0)(0)XN(0)(J(ρ)/ρJ(ρ))t21/3Γ(ρ)2/3A(ρ)1t2/3=du^.\lim_{t\to\infty}\frac{X_{N(t\downarrow0)}(0)-X_N(0)-(J(\rho)/\rho-J'(\rho))t}{2^{1/3}\Gamma(\rho)^{2/3}A(\rho)^{-1}t^{2/3}}\stackrel{d}{=}\hat u.

This predicts the universal argmax distribution of the Airy-two process minus a parabola for backward-geodesic endpoints. The source motivates the scale using KPZ theory and the known TASEP result, but gives no general proof.

Sources & referencesView supporting material

Primary source

Patrik L. Ferrari and Min Liu, “Quasi-geodesics in integrable and non-integrable exclusion processes”, arXiv:2412.11626 (2024).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.