The permanent of a matrix is a calculation similar to the determinant, but without the alternating plus and minus signs. You just sum up products of entries, one from each row and column, over all possible ways to pair up rows and columns. While determinants are easy to compute and well understood, permanents are notoriously difficult, both computationally and theoretically. This paper studies what happens when you fill an n-by-n matrix with entries chosen randomly from a finite field, which is essentially a number system where arithmetic wraps around modulo some prime. The question is: what value does the permanent take, and how is that value distributed?
The conjecture being proved here says that as the matrix size n grows, the permanent becomes essentially equally likely to be any element of the finite field. In other words, even though the permanent is a complicated algebraic function of the entries, its output looks completely uniform, with each possible value occurring with probability roughly 1 divided by the size of the field. A previous conjecture focused just on the probability of the permanent being zero, but the full result says the entire distribution flattens out to uniform. The paper proves this and also gives a quantitative rate, showing the deviation from uniformity shrinks like the logarithm of n divided by n, and this bound works simultaneously for all odd finite fields, not just a fixed one.
The significance of this result is that it confirms a clean and somewhat surprising phenomenon: despite the permanent being a hard-to-control combinatorial sum, random matrices over finite fields produce a permanent that is asymptotically indistinguishable from a uniformly random field element. The proof achieves a bound that is uniform across all odd prime powers, meaning it also handles the case where the field size grows with the matrix size. This places the result in the broader tradition of work showing that natural algebraic statistics of random discrete structures converge to simple limiting distributions, and it resolves a problem that had been open since it was formulated by Ghasemi, Gross, and Kopparty.