Skip to content

Home > Resources > papers > Sch80-Schwartz-Zippel

Fast Probabilistic Algorithms for Verification of Polynomial Identities

Summary

Introduces the probabilistic polynomial identity test: evaluate both sides of a purported identity at a uniformly random point over a large field. If the polynomials are distinct, they can agree at at most d points out of |S| for total degree d, so a random test fails with probability ≤ d/|S|. This bound — now universally called the Schwartz-Zippel lemma — is the information-theoretic engine behind the soundness proofs of virtually every polynomial-based interactive proof, from the sumcheck protocol through PLONK and FRI. An independent, slightly earlier formulation appears in Zippel (1979, EUROSAM) for sparse polynomials; the two results are usually cited together.

Used by