English

Hard properties with (very) short PCPPs and their applications

Computational Complexity 2019-11-18 v2 Discrete Mathematics

Abstract

We show that there exist properties that are maximally hard for testing, while still admitting PCPPs with a proof size very close to linear. Specifically, for every fixed \ell, we construct a property P(){0,1}n\mathcal{P}^{(\ell)}\subseteq\{0,1\}^n satisfying the following: Any testing algorithm for P()\mathcal{P}^{(\ell)} requires Ω(n)\Omega(n) many queries, and yet P()\mathcal{P}^{(\ell)} has a constant query PCPP whose proof size is O(nlog()n)O(n\cdot \log^{(\ell)}n), where log()\log^{(\ell)} denotes the \ell times iterated log function (e.g., log(2)n=loglogn\log^{(2)}n = \log \log n). The best previously known upper bound on the PCPP proof size for a maximally hard to test property was O(npolylogn)O(n \cdot \mathrm{poly}\log{n}). As an immediate application, we obtain stronger separations between the standard testing model and both the tolerant testing model and the erasure-resilient testing model: for every fixed \ell, we construct a property that has a constant-query tester, but requires Ω(n/log()(n))\Omega(n/\log^{(\ell)}(n)) queries for every tolerant or erasure-resilient tester.

Keywords

Cite

@article{arxiv.1909.03255,
  title  = {Hard properties with (very) short PCPPs and their applications},
  author = {Omri Ben-Eliezer and Eldar Fischer and Amit Levi and Ron D. Rothblum},
  journal= {arXiv preprint arXiv:1909.03255},
  year   = {2019}
}
R2 v1 2026-06-23T11:08:31.968Z