arXiv Science⌕ Search
Home/Archive/arXiv

arXiv · 2610.27457

A first order phase transition for lower tails in sparse random graphs

Byron Chin
View original source ↗Read source document ↗Export metadata ↓

Abstract

We study the graphon variational problem that governs lower tail large deviations of the number of copies of a fixed nonbipartite graph $H$ in a sparse Erdős--Rényi random graph. Our main new insight is that any non-constant minimizer of the variational problem exhibits a discontinuous jump away from the constant graphon. We deduce that the replica-symmetric and symmetry-breaking regimes are separated by a single critical parameter and that the phase transition is first order. In the supercritical regime $p \gg n^{-1/m_2(H)}$, we transfer these results to the conditioned random graph, showing that the edge density is a discontinuous order parameter and that the conditioned graph concentrates in cut distance near the full minimizer set.

Explore related subjects

math.PRmath.CO

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Byron Chin. 2026-10-20. A first order phase transition for lower tails in sparse random graphs. https://arxiv.org/abs/2610.27457

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Well-posedness of kinetic McKean-Vlasov equations

We consider the McKean-Vlasov equation $dX_t = b(t, X_t, [X_t])dt + σ(t, X_t, [X_t])dW_t$ where $[X_t]$ is the law of $X_t$. We specifically consider the kinetic case, where the equation is degenerate because the dimension of the Brownian motion $W$ is strictly smaller than that of the solution $X$, as commonly required in classical models of collisional kinetic theory. Assuming Hölder continuous coefficients and a weak Hörmander condition, we prove the well-posedness of the equation. This result advances the existing literature by filling a crucial gap: it addresses the previously unexplored case where the diffusion coefficient $σ$ depends on the law $[X_t]$. Notably, our proof employs a simplified and direct argument eliminating the need for PDEs involving derivatives with respect to the measure argument. A critical ingredient is the sub-Riemannian metric structure induced by the corresponding Fokker-Planck operator.

math.PR↗
arXiv2026

Ergodic properties of the harmonic process

In this paper we study detailed fluctuation results for a class of non-equilibrium steady states. The main example is the boundary driven harmonic model \cite{frassek2022exact}. In this model, the non-equilibrium steady state (NESS) is a mixture of products of geometric distributions, of which the local parameters are in turn distributed as uniform order statistics. For such a NESS, we prove law of large numbers, central limit theorem and large deviation results for fields of a general local functions (generalizing the density field). We also obtain quantitative results on the deviation from local equilibrium.

math.PR↗
arXiv2026

First pair coalescence among three geodesics

In directed last-passage percolation with i.i.d. exponential vertex weights, the coalescence time $\operatorname{Coal}(a,b)$ of two semi-infinite geodesics started from vertices $a$ and $b$ in a fixed direction has a heavy tail with (KPZ) exponent $2/3$, and hence infinite expectation. We study instead the first coalescence among triples of semi-infinite geodesics in first- and last-passage percolation models (FPP/LPP) on $\mathbb{Z}^2$ that arise from stationary compatible families, such as Busemann geodesics. For directed FPP and LPP with arbitrary weights, we prove that $$\mathbb{E}\bigl[\min\{\operatorname{Coal}(a,a-e_1),\operatorname{Coal}(a,a-e_2)\}\bigr]<\infty.$$ We also consider FPP and LPP models in which antidiagonal motion is allowed, and prove that for every integer $k>0$, \[\mathbb{E}\bigl[\min\{\operatorname{Coal}(a,a-k(e_1-e_2)),\operatorname{Coal}(a,a+k(e_1-e_2))\}\bigr]<\infty.\] These results support the heuristic prediction that adjacent coalescence times are negatively associated.

math.PR↗