Gibbsian representation for point processes via hyperedge potentials (Q2224968)

In this perfect paper, the authors are interested in the Euclidean space \(\mathbb{R}^{d}\) with \(d\geq 1\) equipped with its Borel-\(\sigma\)-algebra. The authors setup Gibbsian point processes via Poisson specifications along the lines of \(\lambda\)-specifications for models on fixed geometries as in literature. They give a definition of Gibbs point processes via the Dobrushin-Lanford-Ruelle (DLR) equation. The notion of hyperedge potentials is discussed. Under given conditions, the authors prove that there exists a unique vacuum potential \(\Phi\) on the hypergraph structure \(\mathcal{\overline{E}}^k\) which represents \(\rho\). Moreover, if \(\rho\) is vacuum uniformly log-quasilocal, then the potential \(\Phi\) is uniformly convergent. Let \(\rho\) be a vacuum uniformly log-quasilocal \(\Omega^{*}\)-pre-modification. The authors prove that there exists a representation of \(\rho\) via an absolutely summable potential with uniform finite-horizon property. Also, under given conditions, the authors prove that there exists a representation of \(\rho\) via a translation-invariant absolutely summable potential with uniform finite-horizon property. The authors deal with the potentials for a time-evolved Widom-Rowlinson Model (WRM) such as the vacuum potential for the time-evolved WRM. They examine existence of absolutely summable potentials and existence of absolutely summable translation-invariant potentials. The authors explain that in the Gibbsian regimes there is always a natural hyperedge potential. In the last section, the authors provide the proofs.