The KPZ universality class is a large family of random growth processes, including things like bacterial colony growth, coffee stains spreading, and polymers in random environments. These systems are "exactly solvable" in a mathematical sense, meaning their statistical properties can be computed precisely using sophisticated tools. A key tool is something called a Fredholm determinant, which is a kind of infinite-dimensional generalization of an ordinary matrix determinant, and it encodes the probability distributions of interest. This paper develops a unified algebraic framework that converts Fredholm determinant information for eighteen different models across four different physical regimes into explicit, manageable equations.
The central technical achievement is the derivation of what the author calls "diamond equations," a system of compatibility conditions on a lattice graph. Starting from simple "seed" solutions, the author builds a Darboux transformation, a classical technique that generates new solutions from old ones while preserving the underlying equation structure. The key payoff is that this transformation compresses the unwieldy infinite-dimensional Fredholm data into a finite matrix object that satisfies the same structural equations. This means checking whether a complicated model obeys a closed equation reduces to verifying a small number of linear conditions on the model's basic building blocks, which is a dramatic simplification.
The broader significance is that these diamond equations turn out to be equivalent, through a change of variables, to the non-Abelian Hirota-Miwa system, a well-known and central object in classical integrability theory, the mathematical study of exactly solvable systems. This connection reveals that the random growth models of the KPZ class are secretly governed by the same algebraic skeleton as classical integrable systems, building a concrete bridge between modern probability theory and older mathematical physics. The framework also recovers and generalizes results from the author's earlier work, unifying previously scattered results into a single coherent structure.