Scalable verification of quantized neural networks
Formal verification of neural networks is an active topic of research, and recent advances have significantly increased the size of the networks that verification tools can handle. However, most methods are designed for verification of an idealized model of the actual network which works over real a...
Saved in:
Main Authors: | , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2021
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/9074 https://ink.library.smu.edu.sg/context/sis_research/article/10077/viewcontent/16496_Article_Text_19990_1_2_20210518.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-10077 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-100772024-08-01T15:21:36Z Scalable verification of quantized neural networks HENZINGER, Thomas A. LECHNER, Mathias ZIKELIC, Dorde Formal verification of neural networks is an active topic of research, and recent advances have significantly increased the size of the networks that verification tools can handle. However, most methods are designed for verification of an idealized model of the actual network which works over real arithmetic and ignores rounding imprecisions. This idealization is in stark contrast to network quantization, which is a technique that trades numerical precision for computational efficiency and is, therefore, often applied in practice. Neglecting rounding errors of such low-bit quantized neural networks has been shown to lead to wrong conclusions about the network’s correctness. Thus, the desired approach for verifying quantized neural networks would be one that takes these rounding errors into account. In this paper, we show that verifying the bitexact implementation of quantized neural networks with bitvector specifications is PSPACE-hard, even though verifying idealized real-valued networks and satisfiability of bit-vector specifications alone are each in NP. Furthermore, we explore several practical heuristics toward closing the complexity gap between idealized and bit-exact verification. In particular, we propose three techniques for making SMT-based verification of quantized neural networks more scalable. Our experiments demonstrate that our proposed methods allow a speedup of up to three orders of magnitude over existing approaches. 2021-02-01T08:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/9074 info:doi/10.1609/aaai.v35i5.16496 https://ink.library.smu.edu.sg/context/sis_research/article/10077/viewcontent/16496_Article_Text_19990_1_2_20210518.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Artificial Intelligence and Robotics OS and Networks |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Artificial Intelligence and Robotics OS and Networks |
spellingShingle |
Artificial Intelligence and Robotics OS and Networks HENZINGER, Thomas A. LECHNER, Mathias ZIKELIC, Dorde Scalable verification of quantized neural networks |
description |
Formal verification of neural networks is an active topic of research, and recent advances have significantly increased the size of the networks that verification tools can handle. However, most methods are designed for verification of an idealized model of the actual network which works over real arithmetic and ignores rounding imprecisions. This idealization is in stark contrast to network quantization, which is a technique that trades numerical precision for computational efficiency and is, therefore, often applied in practice. Neglecting rounding errors of such low-bit quantized neural networks has been shown to lead to wrong conclusions about the network’s correctness. Thus, the desired approach for verifying quantized neural networks would be one that takes these rounding errors into account. In this paper, we show that verifying the bitexact implementation of quantized neural networks with bitvector specifications is PSPACE-hard, even though verifying idealized real-valued networks and satisfiability of bit-vector specifications alone are each in NP. Furthermore, we explore several practical heuristics toward closing the complexity gap between idealized and bit-exact verification. In particular, we propose three techniques for making SMT-based verification of quantized neural networks more scalable. Our experiments demonstrate that our proposed methods allow a speedup of up to three orders of magnitude over existing approaches. |
format |
text |
author |
HENZINGER, Thomas A. LECHNER, Mathias ZIKELIC, Dorde |
author_facet |
HENZINGER, Thomas A. LECHNER, Mathias ZIKELIC, Dorde |
author_sort |
HENZINGER, Thomas A. |
title |
Scalable verification of quantized neural networks |
title_short |
Scalable verification of quantized neural networks |
title_full |
Scalable verification of quantized neural networks |
title_fullStr |
Scalable verification of quantized neural networks |
title_full_unstemmed |
Scalable verification of quantized neural networks |
title_sort |
scalable verification of quantized neural networks |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2021 |
url |
https://ink.library.smu.edu.sg/sis_research/9074 https://ink.library.smu.edu.sg/context/sis_research/article/10077/viewcontent/16496_Article_Text_19990_1_2_20210518.pdf |
_version_ |
1814047724399493120 |