What question did this study set out to answer?

This research aims to create a framework for executing graph neural network inference while ensuring data privacy and verifiability.

May 15, 20260 citationsOpen Access

View Full Paper

ZK-GSP: Zero-Knowledge Verifiable Graph Neural Network Inference

DBDaniel Baker

Key Points

This research aims to create a framework for executing graph neural network inference while ensuring data privacy and verifiability.
Developed ZK-GSP integrating GSP operators into a RISC-V executor.
Measured proof generation time and cycle count for graph sizes ranging from N=4 to 32.
Used a zero-knowledge proof system to ensure privacy of graph features and topology.
Proof generation and execution confirmed with STARK receipts without revealing any graph data.
Cycle count for operations scales as O(N^2) relative to graph size.
End-to-end implementation successfully performed in Rust programming language.

Abstract

We propose ZK-GSP, a framework for privacy-preserving, verifiably correct graph neuralnetwork (GNN) inference using zero-knowledge STARKs. The key insight is a structuralisomorphism between RISC Zero's zkVM and graph signal processing (GSP) pipelines:both are typed, deterministic data-flow graphs whose execution can be committed to witha succinct, publicly verifiable proof. By implementing GSP operators as instructions insidethe RISC Zero executor, we obtain a system that runs GNN programs on private graphdata and emits a STARK receipt proving correct execution, without revealing the graphtopology, node features, or intermediate embeddings. We demonstrate an end-to-endimplementation in Rust, measure proof generation time and RISC-V cycle count acrossgraph sizes N = 4 to 32, and find that cycle count scales as O(N2), matching thecomplexity of the underlying GCN operator. Code: github.com/brack-6/zk-gsp

问 AI

Bookmark

View Full Paper

问 AI

Bookmark

View Full Paper

ZK-GSP: Zero-Knowledge Verifiable Graph Neural Network Inference

Key Points

Abstract

Cite This Study