Mutlivariate cryptography
by Danny Willems
Interesting reading: Signature of correct computations (sometimes called PST’13, because accepted at TCC13).
Describe a multivariate polynomial commitment scheme based on some observations from KZG10.
I like the multiset representation of a multivariate polynomial, section 2.4. Also, good description of KZG, see 1.3.
Good reminder to see the verification of a SNARK as a “public key” and the proof as a “signature”. The toxic waste can therefore be seen as a private key that we must not keep.
Maybe the idea of using derivate evaluations $\frac{f(x) - f’(a) x}{(x - a)^2}$ is something to dig into. Wondering if the paper can be used for lookup arguments with large tables, using the decomposition of a polynomial $P(X) - P(a) \in \mathbb{F}[X_{1}, …, X_{N}]$ in $\sum_{i = 1}^{N} (X_{i} - a_{i}) q_{i}(X)$
tags: privacy - mutlivariate - cryptography