This is a series of lectures on zero-knowledge (ZK) proofs, conducted at Distributed Lab, in which we explore "from scratch" how modern zk-SNARKs such as Groth16, PlonK, GKR work, including all the mathematical components they rely on. The course covers not only ZK theory itself and its applications, but also the basic mathematics needed to understand ZK and cryptography in general.
Presentation of the client-side ZKML framework based on UltraGroth and R1CS arithmetization.
Overview of the UltraGroth protocol: Groth16 modification that enables lookup checks and sampling randomness.
Research on implementing Bitcoin on-chain public verifiable computation scheme (based on BitVM2 research).
Lecture about implementing Elliptic Curve (EC) operations efficiently. Basics of EC pairing.
What are neural networks. Their applications, basic dense neural network, demonstration of MNIST dataset training.
Applications of Deep Learning to Biometric Analysis problem. Triplet Loss function, contrastive learning, face recognition models.