Paper 2022/1557

Less is more: refinement proofs for probabilistic proofs

Kunming Jiang, New York University, Carnegie Mellon University
Devora Chait-Roth, New York University
Zachary DeStefano, NYU
Michael Walfish, NYU
Thomas Wies, NYU
Abstract

There has been intense interest over the last decade in implementations of _probabilistic proofs_ (IPs, SNARKs, PCPs, and so on): protocols in which an untrusted party proves to a verifier that a given computation was executed properly, possibly in zero knowledge. Nevertheless, implementations still do not scale beyond small computations. A central source of overhead is the _front-end_: translating from the abstract computation to a set of equivalent arithmetic constraints. This paper introduces a general-purpose framework, called Distiller, in which a user translates to constraints not the original computation but an abstracted _specification_ of it. Distiller is the first in this area to perform such transformations in a way that is provably safe. Furthermore, by taking the idea of "encode a check in the constraints" to its literal logical extreme, Distiller exposes many new opportunities for constraint reduction, resulting in cost reductions for benchmark computations of 1.3–50, and in some cases, better asymptotics.

Metadata
Available format(s)
PDF
Category
Applications
Publication info
Published elsewhere. Minor revision. 2023 IEEE Symposium on Security and Privacy (SP)
DOI
10.1109/SP46215.2023.00142
Keywords
probabilistic proofszero knowledgeoutsourced computationrefinement proofsformal methodswidgetsgadgetsR1CS
Contact author(s)
kunmingj @ andrew cmu edu
dc4451 @ nyu edu
zd2131 @ nyu edu
mwalfish @ cs nyu edu
wies @ cs nyu edu
History
2023-08-02: revised
2022-11-09: received
See all versions
Short URL
https://ia.cr/2022/1557
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2022/1557,
      author = {Kunming Jiang and Devora Chait-Roth and Zachary DeStefano and Michael Walfish and Thomas Wies},
      title = {Less is more: refinement proofs for probabilistic proofs},
      howpublished = {Cryptology {ePrint} Archive, Paper 2022/1557},
      year = {2022},
      doi = {10.1109/SP46215.2023.00142},
      url = {https://eprint.iacr.org/2022/1557}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.