國立虎尾科技大學 |

Neural Network Editing : = Algorithms and Applications.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	Neural Network Editing :/
Reminder of title:	Algorithms and Applications.
Author:	Fu, Feisi.
Description:	1 online resource (114 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-11, Section: B.
Contained By:	Dissertations Abstracts International85-11B.
Subject:	Information technology. -
Online resource:	click for full text (PQDT)
ISBN:	9798382747248

Neural Network Editing : = Algorithms and Applications.
Fu, Feisi.

Neural Network Editing :Algorithms and Applications. - 1 online resource (114 pages)

Source: Dissertations Abstracts International, Volume: 85-11, Section: B.

Thesis (Ph.D.)--Boston University, 2024.

Includes bibliographical references

Deep neural networks have demonstrated impressive performance in a wide variety of applications. However, deep neural networks are not perfect. In many cases, additional adjustments, which we call neural network editing, are essential for various objectives. In this thesis, we present three novel methodologies for neural network editing:1. A novel methodology for repairing neural networks. Unlike existing methods that rely on modifying the weights of a neural network which can induce a global change in the function space, our approach applies only a localized change in the function space while still guaranteeing the removal of the buggy behavior. By leveraging the piecewise linear nature of ReLU networks, our approach can efficiently construct a patch network tailored to the linear region where the buggy input resides, which when combined with the original network, provably corrects the behavior on the buggy input.2. A new approach for repairing pretrained neural networks to satisfy global robustness and individual fairness properties. We prove that any counterexample to a global robustness property must exhibit a corresponding large gradient. For ReLU networks, this result allows us to efficiently identify the linear regions that violate a given global robustness property. By formulating and solving a suitable robust convex optimization problem, our approach then computes a minimal weight change that will provably repair these violating linear regions.3. A novel approach for neural network ownership verification based on the notion of latent watermarks. Our approach to neural network ownership verification is to decouple a network's normal operation from its responses to watermarked inputs during ownership verification. The key idea is to train the network such that the watermarks remain dormant unless the owner's secret key is applied to activate it. The secret key is realized as a specific perturbation only known to the owner to the network's parameters. We show that our approach offers a strong defense against backdoor detection, backdoor removal, and surrogate model attacks.Finally, we summarize the proposed methods and discuss future directions and challenges for neural network editing. This includes exploring the difficulties and challenges of editing neural networks for the latest machine learning models, including large language models.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798382747248Subjects--Topical Terms:

559429
Information technology.
Subjects--Index Terms:

Neural networkIndex Terms--Genre/Form:

554714
Electronic books.

Neural Network Editing : = Algorithms and Applications.
LDR:03738ntm a22004097 4500 001 1148386
005 20240924101912.5
006 m o d
007 cr bn ---uuuuu
008 250605s2024 xx obm 000 0 eng d
020 $a 9798382747248
035 $a (MiAaPQ)AAI31236404
035 $a AAI31236404
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Fu, Feisi. $3 1474338
245 1 0 $a Neural Network Editing : $b Algorithms and Applications.
264 0 $c 2024
300 $a 1 online resource (114 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-11, Section: B.
500 $a Advisor: Li, Wenchao.
502 $a Thesis (Ph.D.)--Boston University, 2024.
504 $a Includes bibliographical references
520 $a Deep neural networks have demonstrated impressive performance in a wide variety of applications. However, deep neural networks are not perfect. In many cases, additional adjustments, which we call neural network editing, are essential for various objectives. In this thesis, we present three novel methodologies for neural network editing:1. A novel methodology for repairing neural networks. Unlike existing methods that rely on modifying the weights of a neural network which can induce a global change in the function space, our approach applies only a localized change in the function space while still guaranteeing the removal of the buggy behavior. By leveraging the piecewise linear nature of ReLU networks, our approach can efficiently construct a patch network tailored to the linear region where the buggy input resides, which when combined with the original network, provably corrects the behavior on the buggy input.2. A new approach for repairing pretrained neural networks to satisfy global robustness and individual fairness properties. We prove that any counterexample to a global robustness property must exhibit a corresponding large gradient. For ReLU networks, this result allows us to efficiently identify the linear regions that violate a given global robustness property. By formulating and solving a suitable robust convex optimization problem, our approach then computes a minimal weight change that will provably repair these violating linear regions.3. A novel approach for neural network ownership verification based on the notion of latent watermarks. Our approach to neural network ownership verification is to decouple a network's normal operation from its responses to watermarked inputs during ownership verification. The key idea is to train the network such that the watermarks remain dormant unless the owner's secret key is applied to activate it. The secret key is realized as a specific perturbation only known to the owner to the network's parameters. We show that our approach offers a strong defense against backdoor detection, backdoor removal, and surrogate model attacks.Finally, we summarize the proposed methods and discuss future directions and challenges for neural network editing. This includes exploring the difficulties and challenges of editing neural networks for the latest machine learning models, including large language models.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Information technology. $3 559429
650 4 $a Computer engineering. $3 569006
650 4 $a Electrical engineering. $3 596380
653 $a Neural network
653 $a ReLU networks
653 $a Watermarks
653 $a Global robustness
653 $a Buggy input
655 7 $a Electronic books. $2 local $3 554714
690 $a 0544
690 $a 0489
690 $a 0464
690 $a 0800
710 2 $a Boston University. $b Systems Engineering ENG. $3 1183006
710 2 $a ProQuest Information and Learning Co. $3 1178819
773 0 $t Dissertations Abstracts International $g 85-11B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31236404 $z click for full text (PQDT)