Check out the latest episode of Node Guardians to see our cofounder David Wong talk about ZK security and the future of ZK with Sam. In this episode, David discusses the different workflows that can be used for blockchain auditing, the importance of math in auditing, and the scope of work that blockchain auditors typically perform. He also talks about the challenges of working in different contexts, how vulnerabilities are rated, and the usefulness of formal verification.
Cryptographic software is error-prone and failures are catastrophic, therefore formal verification is a powerful and increasingly practical tool for greatly improving the assurances of our cryptographic software. But who verifies the formal verification? Formal verification itself is software too, what is proved, under which assumptions and its relation to the real-world implementation lie beyond the scope of the machine-checked proof. In this post, we want to give some insights into our three-week audit of the Ironwood formalization and some of the common pitfalls that formal verification more broadly can encounter.
In Session 11 of "Proof is in the Pudding," we look at what it takes to prove model inference. We cover transformer computation, sumcheck, GKR, lookup arguments, quantization, and KV caching, then discuss scaling and possible uses for verifiable ML.
We think that Clean is the future of circuit development, so we created a programmer-first introduction to show circuit developers who have never used formal verification how they can build circuits faster, safer and better.
We're excited to share that our programming language, noname, now supports R1CS, making it easier to write zero-knowledge (ZK) circuits and deploy them on Ethereum using SnarkJS. This update introduces an alternative to the common Circom language, with a simple and intuitive syntax inspired by Rust and Golang. In this post, we illustrate how to deploy a noname-based Sudoku circuit on Ethereum, demonstrating core benefits like proving a solution's correctness without revealing it. Dive in to explore how noname could potentially unify the fragmented zkSNARK ecosystem and simplify your circuit writing process!
We're super excited to introduce noname 3.0, our zk programming language inspired by Rust and Golang, now achieving full feature parity with Circom. This update brings native hints, a standard library, debugging features, and a lot more to enhance developer experience. Dive into how hint functions work with an 'unsafe' keyword to balance innovation and security, explore our new stdlib modules, and see how the compiler pipeline visualizer can help you understand the compiling process. Plus, check out our next steps and how you can contribute to shaping noname's future.
We're excited to introduce the preview of noname 2.0, packed with features that make developing advanced ZK circuits easier than ever. This update includes flexible generic-sized arrays, seamless integration with folding schemes for IVC, and an interactive online playground to test and share code. We've also optimized R1CS constraint generation to boost performance. Plus, there are numerous community-driven enhancements and bug fixes that make the language more robust and user-friendly. Dive in to explore the specifics of our journey, learn from the contributions of our vibrant open-source community, and see how noname is evolving into a more versatile tool for developers.
Ever wondered how zkVMs simplify the use of zero-knowledge proofs in coding? We dive into how they let developers focus more on application logic by abstracting complex cryptographic aspects, using familiar languages like Rust or C++. But hold on, it's not all smooth sailing. Despite these benefits, a single bug anywhere in the complex system of compilers, proof systems, or verification can lead to serious security issues. In the post, we break down the zkVM workflow, explore common vulnerabilities at each phase, and highlight the importance of understanding these layers to build more secure, zk-powered applications. Curious about how this all plays out? Let’s unravel it together!
Today we're launching zkao, a product by zkSecurity that makes AI security research work the way fuzzing works: not as a one-shot event, but as something you run continuously until coverage compounds.
What if the final form of software development was just watching code and proof popping up while you sip a drink? Letting AI agents write assembly directly alongside Lean proofs sidesteps the whole compiler-trust problem. With a peek at real EVM 256-bit addition code and its specification, you'll see why the assembly + Lean paradigm is final in both the historical and category theoretic sense.