HyperNova Recursion System Enables Practical Zero-Knowledge Virtual Machines
HyperNova, a novel recursive proof system, drastically reduces overhead for high-degree constraint computations, making efficient zkVMs a reality.
Post-Quantum zk-SNARKs from LWE Secure Verifiable Computation for All Circuits
This research formalizes quantum-safe zk-SNARKs for arithmetic circuits using LWE, securing blockchain's verifiable computation layer.
Lattice-Based Inner Product Argument Unlocks Post-Quantum Transparent SNARKs
The Lattice-IPA primitive achieves a succinct, transparent, and quantum-resistant proof system, fundamentally securing verifiable computation against future quantum adversaries.
Zero-Knowledge Bag Unlocks Constant-Time Verifiable General Computation
Introducing the Zero-Knowledge Bag, a new cryptographic primitive enabling constant computational and communication complexity for zkVM execution.
Equifficient Polynomial Commitments Achieve Smallest Proof Size and Fastest SNARKs
Equifficient Polynomial Commitments are a new primitive that enforces polynomial basis representation, enabling SNARKs with 160-byte proofs and triple-speed proving.
Equifficient Polynomial Commitments Enable Ultra-Succinct, Faster Zero-Knowledge Proofs
Equifficient Polynomial Commitments introduce a new cryptographic primitive that separates linear and nonlinear constraints, setting the new frontier for zk-SNARK efficiency.
Equifficient Polynomial Commitments Enable Faster, Smaller zk-SNARKs
Research introduces Equifficient Polynomial Commitments, a new primitive that yields Pari, the smallest SNARK at 160 bytes, and Garuda, a prover three times faster than Groth16.
Linear Prover Time Unlocks Optimal Verifiable Computation Scaling
Introducing FoldCommit, a new polynomial commitment scheme that achieves optimal linear-time prover complexity, fundamentally lowering the cost of generating large-scale zero-knowledge proofs.
ZK Proof of Training Secures Private Decentralized AI Consensus
ZK Proof of Training (ZKPoT) leverages zk-SNARKs to validate model contributions by accuracy, enabling private, scalable, and fair decentralized AI networks.
