Testing-based Black-box Extraction of Simple Models from RNNs and Transformers

被引：0

作者：

Muskardin, Edi ^{[1
,2
]}

Tappler, Martin ^{[1
,2
]}

Aichernig, Bernhard K. ^{[2
]}

机构：

[1] Graz Univ Technol, Silicon Austria Labs, SAL DES Lab, Graz, Austria

[2] Graz Univ Technol, Graz, Austria

来源：

INTERNATIONAL CONFERENCE ON GRAMMATICAL INFERENCE, VOL 217 | 2023年 / 217卷

关键词：

Model extraction; Active automata learning; RNN; Transformers;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this technical report, we outline the testing-based black-box method used to extract simple and interpretable models from RNNs and transformers. Our work was done in the scope of the TAYSIR competition, in which it won the first place.

引用

页码：291 / 294

页数：4

共 50 条

[41] ONEMAX in Black-Box Models with Several Restrictions
Doerr, Carola
Lengler, Johannes
ALGORITHMICA, 2017, 78 (02) : 610 - 640
[42] Auditing black-box models for indirect influence
Adler, Philip
Falk, Casey
Friedler, Sorelle A.
Nix, Tionney
Rybeck, Gabriel
Scheidegger, Carlos
Smith, Brandon
Venkatasubramanian, Suresh
KNOWLEDGE AND INFORMATION SYSTEMS, 2018, 54 (01) : 95 - 122
[43] Demystifying Black-box Models with Symbolic Metamodels
Alaa, Ahmed M.
van der Schaar, Mihaela
ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 32 (NIPS 2019), 2019, 32
[44] Auditing Black-box Models for Indirect Influence
Adler, Philip
Falk, Casey
Friedler, Sorelle A.
Rybeck, Gabriel
Scheidegger, Carlos
Smith, Brandon
Venkatasubramanian, Suresh
2016 IEEE 16TH INTERNATIONAL CONFERENCE ON DATA MINING (ICDM), 2016, : 1 - 10
[45] Regularizing Black-box Models for Improved Interpretability
Plumb, Gregory
Al-Shedivat, Maruan
Cabrera, Angel Alexander
Perer, Adam
Xing, Eric
Talwalkar, Ameet
ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020, 2020, 33
[46] BLACK-BOX MODELS FOR LINEAR INTEGRATED CIRCUITS
MURRAYLA.MA
IEEE TRANSACTIONS ON EDUCATION, 1969, E 12 (03) : 170 - &
[47] Learning Groupwise Explanations for Black-Box Models
Gao, Jingyue
Wang, Xiting
Wang, Yasha
Yan, Yulan
Xie, Xing
PROCEEDINGS OF THE THIRTIETH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, IJCAI 2021, 2021, : 2396 - 2402
[48] Explaining Black-box Classification Models with Arguments
Amgoud, Leila
2021 IEEE 33RD INTERNATIONAL CONFERENCE ON TOOLS WITH ARTIFICIAL INTELLIGENCE (ICTAI 2021), 2021, : 791 - 795
[49] Adversarial Eigen Attack on Black-Box Models
Zhou, Linjun
Cui, Peng
Zhang, Xingxuan
Jiang, Yinan
Yang, Shiqiang
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022), 2022, : 15233 - 15241
[50] Auditing black-box models for indirect influence
Philip Adler
Casey Falk
Sorelle A. Friedler
Tionney Nix
Gabriel Rybeck
Carlos Scheidegger
Brandon Smith
Suresh Venkatasubramanian
Knowledge and Information Systems, 2018, 54 : 95 - 122

← 1 2 3 4 5 →