Towards Efficient Verification of Quantized Neural Networks
Authors
- Pei HuangStanford University, Stanford, CA, USA
- Haoze WuStanford University, Stanford, CA, USA
- Yuting YangInstitute of Computing Technology, Chinese Academy of Sciences, Beijing, China
- Ieva DaukantasIT University of Copenhagen, Copenhagen, Denmark
- Min WuStanford University, Stanford, CA, USA
- Yedi ZhangNational University of Singapore, Singapore
- Clark BarrettStanford University, Stanford, CA, USA
DOI:
https://doi.org/10.1609/aaai.v38i19.30108Keywords:
GeneralAbstract
Quantization replaces floating point arithmetic with integer arithmetic in deep neural network models, providing more efficient on-device inference with less power and memory. In this work, we propose a framework for formally verifying the properties of quantized neural networks. Our baseline technique is based on integer linear programming which guarantees both soundness and completeness. We then show how efficiency can be improved by utilizing gradient-based heuristic search methods and also bound-propagation techniques. We evaluate our approach on perception networks quantized with PyTorch. Our results show that we can verify quantized networks with better scalability and efficiency than the previous state of the art.
Downloads
Published
2024-03-24
How to Cite
Huang, P., Wu, H., Yang, Y., Daukantas, I., Wu, M., Zhang, Y., & Barrett, C. (2024). Towards Efficient Verification of Quantized Neural Networks.Proceedings of the AAAI Conference on Artificial Intelligence,38(19), 21152-21160. https://doi.org/10.1609/aaai.v38i19.30108
Issue
Section
AAAI Technical Track on Safe, Robust and Responsible AI Track
