Tag
This paper presents the first systematic study of zero-knowledge-friendly quantization for LLMs, formalizing the concept and evaluating nine models (including Qwen2.5-14B and Qwen3-30B-A3B) across weight, activation, and nonlinear lookup table precision. It finds that activation precision and RMSNorm inverse-square-root lookups dominate ZK proving costs and that conventional low-bit quantization heuristics do not translate to proportional proving savings.
Signal introduces zero-knowledge proofs to allow user registration without requiring a phone number, enhancing privacy and security.
The article argues that Zero-Knowledge Proofs are not an ideal solution for age verification, highlighting risks such as privacy breaches and centralization of internet access control.
PANDA is a scalable system using zero-knowledge proofs to verify the robustness and fairness of neural networks without revealing model parameters, enabling certification for large networks with polynomial complexity.
This article explains zero-knowledge proofs by focusing on a protocol for graph 3-coloring from the Goldreich-Micali-Widgerson paper and shares a simple implementation.
zkGolf is an open competition to produce optimized, formally verified zero-knowledge circuits, using LLMs to generate and improve implementations beyond human-optimized standards.
This paper introduces directional sharpness, a new metric for certifying the generalization performance of machine learning models that is both efficient to compute and more reliable than existing proxies like test accuracy or traditional sharpness, even when training deviates from prescribed procedures.
Former Signal developers and cryptographers from Harvard and Microsoft Research release Encrypted Spaces, a set of open-source libraries enabling developers to build end-to-end encrypted collaborative apps using zero-knowledge proofs.
A new type of zero-knowledge proof leverages Gödel's incompleteness theorems to overcome previous limitations of secrecy, establishing a striking connection between mathematical logic and cryptography.