Show simple item record

dc.contributor.author
Mirman, Matthew
dc.contributor.author
Singh, Gagandeep
dc.contributor.author
Vechev, Martin
dc.date.accessioned
2020-01-29T12:43:57Z
dc.date.available
2020-01-29T12:10:40Z
dc.date.available
2020-01-29T12:43:57Z
dc.date.issued
2019-03
dc.identifier.uri
http://hdl.handle.net/20.500.11850/395401
dc.identifier.doi
10.3929/ethz-b-000395401
dc.description.abstract
We present a training system, which can provably defend significantly larger neural networks than previously possible, including ResNet-34 and DenseNet-100. Our approach is based on differentiable abstract interpretation and introduces two novel concepts: (i) abstract layers for fine-tuning the precision and scalability of the abstraction, (ii) a flexible domain specific language (DSL) for describing training objectives that combine abstract and concrete losses with arbitrary specifications. Our training method is implemented in the DiffAI system.
en_US
dc.format
application/pdf
en_US
dc.language.iso
en
en_US
dc.publisher
Cornell University
en_US
dc.rights.uri
http://rightsstatements.org/page/InC-NC/1.0/
dc.title
A Provable Defense for Deep Residual Networks
en_US
dc.type
Working Paper
dc.rights.license
In Copyright - Non-Commercial Use Permitted
ethz.journal.title
arXiv
ethz.pages.start
1903.12519
en_US
ethz.size
30 p.
en_US
ethz.identifier.arxiv
1903.12519
ethz.publication.place
Ithaca, NY
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02150 - Dep. Informatik / Dep. of Computer Science::02664 - Inst. f. Programmiersprachen u. -systeme / Inst. Programming Languages and Systems::03948 - Vechev, Martin / Vechev, Martin
en_US
ethz.leitzahl.certified
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02150 - Dep. Informatik / Dep. of Computer Science::02664 - Inst. f. Programmiersprachen u. -systeme / Inst. Programming Languages and Systems::03948 - Vechev, Martin / Vechev, Martin
en_US
ethz.tag
A Provable Defense for Deep Residual Networks
en_US
ethz.date.deposited
2020-01-29T12:10:48Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Open access
en_US
ethz.rosetta.installDate
2020-01-29T12:44:07Z
ethz.rosetta.lastUpdated
2020-02-15T23:55:08Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=A%20Provable%20Defense%20for%20Deep%20Residual%20Networks&rft.jtitle=arXiv&rft.date=2019-03&rft.spage=1903.12519&rft.au=Mirman,%20Matthew&Singh,%20Gagandeep&Vechev,%20Martin&rft.genre=preprint&
 Search print copy at ETH Library

Files in this item

Thumbnail

Publication type

Show simple item record