Highlights

Understanding Binius

两篇文章介绍了 Ligero-Binius 和 FRI-Binius 零知识证明系统，探讨其多项式承诺方案、高效性、递归验证优化及硬件加速潜力。

The two articles introduce Ligero-Binius and FRI-Binius zero-knowledge proof systems, discussing their polynomial commitment schemes, efficiency improvements, recursive verification optimizations, and potential for hardware acceleration.

• https://hackmd.io/@l2iterative/binius
• https://hackmd.io/@l2iterative/binius2

Terence Tao: Machine-Assisted Proof

文章探讨计算机辅助数学研究，包括机器学习、形式化证明助手、大语言模型的应用及其潜力与挑战。

The article explores computer-assisted mathematical research, covering applications of machine learning, formal proof assistants, and large language models, highlighting their current potential, challenges, and integration for future advancements in mathematics.

• https://www.ams.org/notices/202501/rnoti-p6.pdf

Nethermind: Introducing LatticeFold Rust implementation

由 Nethermind 基于 Dan Boneh 和 Binyi Chen 的研究 LatticeFold: A Lattice-based Folding Scheme and its Applications to Succinct Proof Systems 开发的 LatticeFold 折叠方案的概念验证实现。

A proof-of-concept implementation of the LatticeFold folding scheme engineered by Nethermind based on the work LatticeFold: A Lattice-based Folding Scheme and its Applications to Succinct Proof Systems by Dan Boneh and Binyi Chen.

• https://x.com/NethermindEth/status/1872494705839374681
• https://github.com/NethermindEth/latticefold

MyZKP: Building Zero Knowledge Proof from Scratch in Rust

MyZKP 是一个完全从零开始用 Rust 实现的零知识协议项目！该项目旨在作为理解和使用零知识证明的教育资源。

MyZKP is a Rust implementation of zero-knowledge protocols built entirely from scratch! This project serves as an educational resource for understanding and working with zero-knowledge proofs.

• https://koukyosyumei.github.io/MyZKP/index.html

Episode 346: ZK in Review: Decoding 2024 & Predicting 2025

他们讨论生态系统面临的挑战并简要展望了 2025 年 ZK 的发展前景。

They go on to discuss the challenges facing the ecosystem. And they wrap with a brief look forward, covering what to expect in ZK in 2025.

• https://zeroknowledge.fm/346-2/

Hackathon: MHEGA - Make Homomorphic Encryption Great Again

通过调整 HElib 探索了在 coSNARKs 中使用同态加密（HE）。尽管实现了一些优化，但 HE 中的 FFT 仍比普通方法或 MPC 慢 1 万倍，并且需要大量内存。

Explored using HE for coSNARKs by adapting HElib. Despite getting optimizations to work, FFTs in HE are still 10,000 times slower than plain or MPC and need huge memory.

• https://blog.taceo.io/mhega/

Papers

(Deep) Learning about Elliptic Curve Cryptography

• https://eprint.iacr.org/2024/2064

Bypassing the characteristic bound in logUp

• https://eprint.iacr.org/2024/2067

Zero Knowledge Memory-Checking Techniques for Stacks and Queues

• https://eprint.iacr.org/2024/2084

How To Think About End-To-End Encryption and AI: Training, Processing, Disclosure, and Consent

• https://eprint.iacr.org/2024/2086