Institution profile

NEAR Protocol

Industry researchnorthamerica · us
Official website
Research library1linked papers
Opportunities0open roles
Selected work

Representative Papers

CrowdProve: Community Proving for ZK Rollups

Jan 06, 2025

To address the scalability limitations, weakened decentralization, and high verification costs caused by centralized provers in ZK Rollups, this paper proposes the first community-driven, decentralized proof generation paradigm tailored for ZK Rollups. Our approach operates over a heterogeneous, unreliable hardware network and integrates dynamic task sharding, lightweight proof sharding, redundant submission, and multi-replica consistency verification. We further design an incentive-compatible contributor collaboration protocol and hardware-adaptive adaptation techniques. Experimental evaluation in realistic ZK Rollup settings demonstrates that our system achieves proof latency and throughput comparable to—and often exceeding—state-of-the-art centralized provers, while requiring only mid-to-low-end device clusters to deliver equivalent performance. This significantly reduces operational costs without compromising security or decentralization guarantees.

0 citationsRead paper
Recent publications

Latest Papers

CrowdProve: Community Proving for ZK Rollups

Jan 06, 2025

To address the scalability limitations, weakened decentralization, and high verification costs caused by centralized provers in ZK Rollups, this paper proposes the first community-driven, decentralized proof generation paradigm tailored for ZK Rollups. Our approach operates over a heterogeneous, unreliable hardware network and integrates dynamic task sharding, lightweight proof sharding, redundant submission, and multi-replica consistency verification. We further design an incentive-compatible contributor collaboration protocol and hardware-adaptive adaptation techniques. Experimental evaluation in realistic ZK Rollup settings demonstrates that our system achieves proof latency and throughput comparable to—and often exceeding—state-of-the-art centralized provers, while requiring only mid-to-low-end device clusters to deliver equivalent performance. This significantly reduces operational costs without compromising security or decentralization guarantees.

0 citationsRead paper