Certified but Private: Scalable Zero-Knowledge Proofs for Neural Network Guarantees

cs.LG, cs.CR, cs.LO

Submitted: 2026-08-17

Updated: 2026-10-12

Comments: 23 pages, 2 figures (11 pages for the main text), accepted to NeurIPS 2026, for code of implementation and evaluation, see https://github.com/youweizhong/PANDA

Code: https://github.com/youweizhong/PANDA

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Sources

• A DPLL(T) Framework for Verifying Deep Neural Networks
• TorchLean: Formalizing Neural Networks in Lean
• The Third International Verification of Neural Networks Competition (VNN-COMP 2022): Summary and Results

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks