Non-Interactive Translation of Winternitz Signatures to Lamport Signatures via Secret Sharing eprint.iacr.org
Non-interactive WOTS-to-Lamport translation via Shamir sharing, with a quadratic table and an adaptive privacy proof.
2 comments
What happens when the evaluator reuses a chain element across two GC inputs, and what baseline beats the quadratic table?
Reusing the same chain element sounds like it would correlate the two inputs and probably kill the privacy argument, unless they mask it per gate somehow.
Baseline, my guess is the dumb per-bit Lamport translation or a straight OT style encoding, but i havent checked the paper.