Sound Debloating of Redundant Checks in Zero-Knowledge Machine-Learning Circuits
Automated framework soundly removes up to 48.7% of redundant constraints in ezkl and zkml ZK-ML circuits, cutting prover time by up to 72.8%.
The framework uses whole-circuit abstract interpretation and a provenance graph to verify that each removed redundant check (range proofs, sign lookups, bit decompositions) remains entailed by the rest of the circuit, provably preserving soundness. It was evaluated on MLP, CNN, RNN, and transformer circuits generated by ezkl and zkml, with up to 25.3 million constraints. It removes up to 48.7% of constraints and reduces prover time by up to 72.8% without weakening security. Under-constrained circuits in deployed ZK systems have previously enabled attackers to forge transactions and bypass verification.
- Automated debloating provably preserves soundness of ZK-ML circuits
- Removes up to 48.7% of constraints across ezkl and zkml circuits
- Prover time reduced by up to 72.8% on circuits up to 25.3M constraints
- Under-constrained ZK circuits have enabled real forging attacks in deployed systems
From the original article
1 min read at arxiv.orgZero-knowledge (ZK) proof systems for neural-network inference compile the model into a system of arithmetic constraints. Many of these constraints are redundant checks: range proofs, sign lookups, and bit decompositions who are globally entailed by the rest of the circuit through chains of reasoning that span distant gadgets.
By Zhantong Xue · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.