A Survey on Knowledge Editing of Neural Networks

Deep neural networks are becoming increasingly pervasive in academia and industry, matching and surpassing human performance on a wide variety of fields and related tasks. However, just as humans, even the largest artificial neural networks make mistakes, and once-correct predictions can become inva...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2024-08
Hauptverfasser: Mazzia, Vittorio, Pedrani, Alessandro, Caciolai, Andrea, Rottmann, Kay, Bernardi, Davide
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Mazzia, Vittorio
Pedrani, Alessandro
Caciolai, Andrea
Rottmann, Kay
Bernardi, Davide
description Deep neural networks are becoming increasingly pervasive in academia and industry, matching and surpassing human performance on a wide variety of fields and related tasks. However, just as humans, even the largest artificial neural networks make mistakes, and once-correct predictions can become invalid as the world progresses in time. Augmenting datasets with samples that account for mistakes or up-to-date information has become a common workaround in practical applications. However, the well-known phenomenon of catastrophic forgetting poses a challenge in achieving precise changes in the implicitly memorized knowledge of neural network parameters, often requiring a full model re-training to achieve desired behaviors. That is expensive, unreliable, and incompatible with the current trend of large self-supervised pre-training, making it necessary to find more efficient and effective methods for adapting neural network models to changing data. To address this need, knowledge editing is emerging as a novel area of research that aims to enable reliable, data-efficient, and fast changes to a pre-trained target model, without affecting model behaviors on previously learned tasks. In this survey, we provide a brief review of this recent artificial intelligence field of research. We first introduce the problem of editing neural networks, formalize it in a common framework and differentiate it from more notorious branches of research such as continuous learning. Next, we provide a review of the most relevant knowledge editing approaches and datasets proposed so far, grouping works under four different families: regularization techniques, meta-learning, direct model editing, and architectural strategies. Finally, we outline some intersections with other fields of research and potential directions for future works.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2884478693</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2884478693</sourcerecordid><originalsourceid>FETCH-proquest_journals_28844786933</originalsourceid><addsrcrecordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mQwcFQILi0qS61UyM9T8M7LL89JTUlPVXBNySzJzEtXyE9T8EstLUrMAVIl5flF2cU8DKxpiTnFqbxQmptB2c01xNlDt6Aov7A0tbgkPiu_tCgPKBVvZGFhYmJuYWZpbEycKgAmPTK0</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2884478693</pqid></control><display><type>article</type><title>A Survey on Knowledge Editing of Neural Networks</title><source>Free E- Journals</source><creator>Mazzia, Vittorio ; Pedrani, Alessandro ; Caciolai, Andrea ; Rottmann, Kay ; Bernardi, Davide</creator><creatorcontrib>Mazzia, Vittorio ; Pedrani, Alessandro ; Caciolai, Andrea ; Rottmann, Kay ; Bernardi, Davide</creatorcontrib><description>Deep neural networks are becoming increasingly pervasive in academia and industry, matching and surpassing human performance on a wide variety of fields and related tasks. However, just as humans, even the largest artificial neural networks make mistakes, and once-correct predictions can become invalid as the world progresses in time. Augmenting datasets with samples that account for mistakes or up-to-date information has become a common workaround in practical applications. However, the well-known phenomenon of catastrophic forgetting poses a challenge in achieving precise changes in the implicitly memorized knowledge of neural network parameters, often requiring a full model re-training to achieve desired behaviors. That is expensive, unreliable, and incompatible with the current trend of large self-supervised pre-training, making it necessary to find more efficient and effective methods for adapting neural network models to changing data. To address this need, knowledge editing is emerging as a novel area of research that aims to enable reliable, data-efficient, and fast changes to a pre-trained target model, without affecting model behaviors on previously learned tasks. In this survey, we provide a brief review of this recent artificial intelligence field of research. We first introduce the problem of editing neural networks, formalize it in a common framework and differentiate it from more notorious branches of research such as continuous learning. Next, we provide a review of the most relevant knowledge editing approaches and datasets proposed so far, grouping works under four different families: regularization techniques, meta-learning, direct model editing, and architectural strategies. Finally, we outline some intersections with other fields of research and potential directions for future works.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Artificial intelligence ; Artificial neural networks ; Datasets ; Editing ; Human performance ; Machine learning ; Neural networks ; Regularization</subject><ispartof>arXiv.org, 2024-08</ispartof><rights>2024. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>780,784</link.rule.ids></links><search><creatorcontrib>Mazzia, Vittorio</creatorcontrib><creatorcontrib>Pedrani, Alessandro</creatorcontrib><creatorcontrib>Caciolai, Andrea</creatorcontrib><creatorcontrib>Rottmann, Kay</creatorcontrib><creatorcontrib>Bernardi, Davide</creatorcontrib><title>A Survey on Knowledge Editing of Neural Networks</title><title>arXiv.org</title><description>Deep neural networks are becoming increasingly pervasive in academia and industry, matching and surpassing human performance on a wide variety of fields and related tasks. However, just as humans, even the largest artificial neural networks make mistakes, and once-correct predictions can become invalid as the world progresses in time. Augmenting datasets with samples that account for mistakes or up-to-date information has become a common workaround in practical applications. However, the well-known phenomenon of catastrophic forgetting poses a challenge in achieving precise changes in the implicitly memorized knowledge of neural network parameters, often requiring a full model re-training to achieve desired behaviors. That is expensive, unreliable, and incompatible with the current trend of large self-supervised pre-training, making it necessary to find more efficient and effective methods for adapting neural network models to changing data. To address this need, knowledge editing is emerging as a novel area of research that aims to enable reliable, data-efficient, and fast changes to a pre-trained target model, without affecting model behaviors on previously learned tasks. In this survey, we provide a brief review of this recent artificial intelligence field of research. We first introduce the problem of editing neural networks, formalize it in a common framework and differentiate it from more notorious branches of research such as continuous learning. Next, we provide a review of the most relevant knowledge editing approaches and datasets proposed so far, grouping works under four different families: regularization techniques, meta-learning, direct model editing, and architectural strategies. Finally, we outline some intersections with other fields of research and potential directions for future works.</description><subject>Artificial intelligence</subject><subject>Artificial neural networks</subject><subject>Datasets</subject><subject>Editing</subject><subject>Human performance</subject><subject>Machine learning</subject><subject>Neural networks</subject><subject>Regularization</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mQwcFQILi0qS61UyM9T8M7LL89JTUlPVXBNySzJzEtXyE9T8EstLUrMAVIl5flF2cU8DKxpiTnFqbxQmptB2c01xNlDt6Aov7A0tbgkPiu_tCgPKBVvZGFhYmJuYWZpbEycKgAmPTK0</recordid><startdate>20240830</startdate><enddate>20240830</enddate><creator>Mazzia, Vittorio</creator><creator>Pedrani, Alessandro</creator><creator>Caciolai, Andrea</creator><creator>Rottmann, Kay</creator><creator>Bernardi, Davide</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20240830</creationdate><title>A Survey on Knowledge Editing of Neural Networks</title><author>Mazzia, Vittorio ; Pedrani, Alessandro ; Caciolai, Andrea ; Rottmann, Kay ; Bernardi, Davide</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_28844786933</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Artificial intelligence</topic><topic>Artificial neural networks</topic><topic>Datasets</topic><topic>Editing</topic><topic>Human performance</topic><topic>Machine learning</topic><topic>Neural networks</topic><topic>Regularization</topic><toplevel>online_resources</toplevel><creatorcontrib>Mazzia, Vittorio</creatorcontrib><creatorcontrib>Pedrani, Alessandro</creatorcontrib><creatorcontrib>Caciolai, Andrea</creatorcontrib><creatorcontrib>Rottmann, Kay</creatorcontrib><creatorcontrib>Bernardi, Davide</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Mazzia, Vittorio</au><au>Pedrani, Alessandro</au><au>Caciolai, Andrea</au><au>Rottmann, Kay</au><au>Bernardi, Davide</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>A Survey on Knowledge Editing of Neural Networks</atitle><jtitle>arXiv.org</jtitle><date>2024-08-30</date><risdate>2024</risdate><eissn>2331-8422</eissn><abstract>Deep neural networks are becoming increasingly pervasive in academia and industry, matching and surpassing human performance on a wide variety of fields and related tasks. However, just as humans, even the largest artificial neural networks make mistakes, and once-correct predictions can become invalid as the world progresses in time. Augmenting datasets with samples that account for mistakes or up-to-date information has become a common workaround in practical applications. However, the well-known phenomenon of catastrophic forgetting poses a challenge in achieving precise changes in the implicitly memorized knowledge of neural network parameters, often requiring a full model re-training to achieve desired behaviors. That is expensive, unreliable, and incompatible with the current trend of large self-supervised pre-training, making it necessary to find more efficient and effective methods for adapting neural network models to changing data. To address this need, knowledge editing is emerging as a novel area of research that aims to enable reliable, data-efficient, and fast changes to a pre-trained target model, without affecting model behaviors on previously learned tasks. In this survey, we provide a brief review of this recent artificial intelligence field of research. We first introduce the problem of editing neural networks, formalize it in a common framework and differentiate it from more notorious branches of research such as continuous learning. Next, we provide a review of the most relevant knowledge editing approaches and datasets proposed so far, grouping works under four different families: regularization techniques, meta-learning, direct model editing, and architectural strategies. Finally, we outline some intersections with other fields of research and potential directions for future works.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2024-08
issn 2331-8422
language eng
recordid cdi_proquest_journals_2884478693
source Free E- Journals
subjects Artificial intelligence
Artificial neural networks
Datasets
Editing
Human performance
Machine learning
Neural networks
Regularization
title A Survey on Knowledge Editing of Neural Networks
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-23T15%3A58%3A39IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=A%20Survey%20on%20Knowledge%20Editing%20of%20Neural%20Networks&rft.jtitle=arXiv.org&rft.au=Mazzia,%20Vittorio&rft.date=2024-08-30&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2884478693%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2884478693&rft_id=info:pmid/&rfr_iscdi=true