The paper tackles a classical problem in mathematics: when is a certain symmetric expression involving many variables always non-negative? The expressions in question are built from "orbit sums," which are symmetric combinations of monomials (products of variables raised to powers). Specifically, the paper looks at a combination of three such orbit sums and asks when this combination stays non-negative for all positive values of the variables, regardless of how many variables there are.
The central finding is a dramatic simplification: to check whether such an expression is non-negative across any number of variables and any dimension, you only need to verify it on a two-variable slice, where all but two of the variables are set to one. This is surprising because the full problem involves arbitrarily many variables and seems to grow in complexity with dimension, yet a fixed two-dimensional check completely captures the answer. The authors prove this by constructing a chain of increasingly strong conditions, showing that each condition in the chain implies the next, so that the weakest non-trivial condition (the two-variable check) propagates all the way up to guarantee global non-negativity.
The technical machinery involves combinatorial and algebraic tools, including a hierarchy of conditions indexed by the number of variables active at once, and careful arguments about how swapping and rearranging subsets of variables preserves or improves the inequality. The paper is a companion to an earlier work that handled a specific difficult step in the hierarchy involving three variables; here that step is bypassed using an independent argument, making the proof self-contained. The result also extends cleanly from integer exponents to rational ones, broadening its applicability.