Scrub-and-Learn: Category-Aware Weight Modification for Machine Unlearning Article Swipe
Jiali Wang
,
Hongxia Bie
,
J. Zhao
,
Yichen Zhi
·
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.3390/ai6060108
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.3390/ai6060108
(1) Background: Machine unlearning plays a crucial role in privacy protection and model optimization, particularly in forgetting entire categories of data in classification tasks. However, existing methods often struggle with high computational costs, such as estimating the inverse Hessian, or require access to the original training data, limiting their practicality. (2) Methods: In this work, we introduce Scrub-and-Learn, which is a category-aware weight modification framework designed to remove class-level knowledge efficiently. By modeling unlearning as a continual learning task, our method leverages re-encoded labels of samples from the target category to guide weight updates, effectively scrubbing unwanted knowledge while preserving the rest of the model’s capacity. (3) Results and Conclusions: Experimental results on multiple benchmarks demonstrate that our method effectively eliminates targeted categories—achieving a recognition rate below 5%—while preserving the performance of retained classes within a 4% deviation from the original model.
Related Topics
Metadata
- Type
- article
- Language
- en
- Landing Page
- https://doi.org/10.3390/ai6060108
- https://www.mdpi.com/2673-2688/6/6/108/pdf?version=1747918810
- OA Status
- gold
- References
- 43
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4410597073
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4410597073Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.3390/ai6060108Digital Object Identifier
- Title
-
Scrub-and-Learn: Category-Aware Weight Modification for Machine UnlearningWork title
- Type
-
articleOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2025Year of publication
- Publication date
-
2025-05-22Full publication date if available
- Authors
-
Jiali Wang, Hongxia Bie, J. Zhao, Yichen ZhiList of authors in order
- Landing page
-
https://doi.org/10.3390/ai6060108Publisher landing page
- PDF URL
-
https://www.mdpi.com/2673-2688/6/6/108/pdf?version=1747918810Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
goldOpen access status per OpenAlex
- OA URL
-
https://www.mdpi.com/2673-2688/6/6/108/pdf?version=1747918810Direct OA link when available
- Concepts
-
Computer science, Cognitive science, Artificial intelligence, PsychologyTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
0Total citation count in OpenAlex
- References (count)
-
43Number of works referenced by this work
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4410597073 |
|---|---|
| doi | https://doi.org/10.3390/ai6060108 |
| ids.doi | https://doi.org/10.3390/ai6060108 |
| ids.openalex | https://openalex.org/W4410597073 |
| fwci | 0.0 |
| type | article |
| title | Scrub-and-Learn: Category-Aware Weight Modification for Machine Unlearning |
| biblio.issue | 6 |
| biblio.volume | 6 |
| biblio.last_page | 108 |
| biblio.first_page | 108 |
| topics[0].id | https://openalex.org/T12072 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.9217000007629395 |
| topics[0].domain.id | https://openalex.org/domains/3 |
| topics[0].domain.display_name | Physical Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/1702 |
| topics[0].subfield.display_name | Artificial Intelligence |
| topics[0].display_name | Machine Learning and Algorithms |
| is_xpac | False |
| apc_list.value | 1000 |
| apc_list.currency | CHF |
| apc_list.value_usd | 1082 |
| apc_paid.value | 1000 |
| apc_paid.currency | CHF |
| apc_paid.value_usd | 1082 |
| concepts[0].id | https://openalex.org/C41008148 |
| concepts[0].level | 0 |
| concepts[0].score | 0.4427374303340912 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[0].display_name | Computer science |
| concepts[1].id | https://openalex.org/C188147891 |
| concepts[1].level | 1 |
| concepts[1].score | 0.3312378525733948 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q147638 |
| concepts[1].display_name | Cognitive science |
| concepts[2].id | https://openalex.org/C154945302 |
| concepts[2].level | 1 |
| concepts[2].score | 0.3264840543270111 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q11660 |
| concepts[2].display_name | Artificial intelligence |
| concepts[3].id | https://openalex.org/C15744967 |
| concepts[3].level | 0 |
| concepts[3].score | 0.29338470101356506 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q9418 |
| concepts[3].display_name | Psychology |
| keywords[0].id | https://openalex.org/keywords/computer-science |
| keywords[0].score | 0.4427374303340912 |
| keywords[0].display_name | Computer science |
| keywords[1].id | https://openalex.org/keywords/cognitive-science |
| keywords[1].score | 0.3312378525733948 |
| keywords[1].display_name | Cognitive science |
| keywords[2].id | https://openalex.org/keywords/artificial-intelligence |
| keywords[2].score | 0.3264840543270111 |
| keywords[2].display_name | Artificial intelligence |
| keywords[3].id | https://openalex.org/keywords/psychology |
| keywords[3].score | 0.29338470101356506 |
| keywords[3].display_name | Psychology |
| language | en |
| locations[0].id | doi:10.3390/ai6060108 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4210200511 |
| locations[0].source.issn | 2673-2688 |
| locations[0].source.type | journal |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | 2673-2688 |
| locations[0].source.is_core | True |
| locations[0].source.is_in_doaj | True |
| locations[0].source.display_name | AI |
| locations[0].source.host_organization | https://openalex.org/P4310310987 |
| locations[0].source.host_organization_name | Multidisciplinary Digital Publishing Institute |
| locations[0].source.host_organization_lineage | https://openalex.org/P4310310987 |
| locations[0].source.host_organization_lineage_names | Multidisciplinary Digital Publishing Institute |
| locations[0].license | cc-by |
| locations[0].pdf_url | https://www.mdpi.com/2673-2688/6/6/108/pdf?version=1747918810 |
| locations[0].version | publishedVersion |
| locations[0].raw_type | journal-article |
| locations[0].license_id | https://openalex.org/licenses/cc-by |
| locations[0].is_accepted | True |
| locations[0].is_published | True |
| locations[0].raw_source_name | AI |
| locations[0].landing_page_url | https://doi.org/10.3390/ai6060108 |
| locations[1].id | pmh:oai:doaj.org/article:58771bf783444c149962ac980404c09a |
| locations[1].is_oa | False |
| locations[1].source.id | https://openalex.org/S4306401280 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | False |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | DOAJ (DOAJ: Directory of Open Access Journals) |
| locations[1].source.host_organization | |
| locations[1].source.host_organization_name | |
| locations[1].license | |
| locations[1].pdf_url | |
| locations[1].version | submittedVersion |
| locations[1].raw_type | article |
| locations[1].license_id | |
| locations[1].is_accepted | False |
| locations[1].is_published | False |
| locations[1].raw_source_name | AI, Vol 6, Iss 6, p 108 (2025) |
| locations[1].landing_page_url | https://doaj.org/article/58771bf783444c149962ac980404c09a |
| indexed_in | crossref, doaj |
| authorships[0].author.id | https://openalex.org/A5076261285 |
| authorships[0].author.orcid | https://orcid.org/0000-0002-8916-4372 |
| authorships[0].author.display_name | Jiali Wang |
| authorships[0].countries | CN |
| authorships[0].affiliations[0].institution_ids | https://openalex.org/I139759216 |
| authorships[0].affiliations[0].raw_affiliation_string | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| authorships[0].institutions[0].id | https://openalex.org/I139759216 |
| authorships[0].institutions[0].ror | https://ror.org/04w9fbh59 |
| authorships[0].institutions[0].type | education |
| authorships[0].institutions[0].lineage | https://openalex.org/I139759216 |
| authorships[0].institutions[0].country_code | CN |
| authorships[0].institutions[0].display_name | Beijing University of Posts and Telecommunications |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Jiali Wang |
| authorships[0].is_corresponding | False |
| authorships[0].raw_affiliation_strings | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| authorships[1].author.id | https://openalex.org/A5060471213 |
| authorships[1].author.orcid | https://orcid.org/0000-0002-3675-8476 |
| authorships[1].author.display_name | Hongxia Bie |
| authorships[1].countries | CN |
| authorships[1].affiliations[0].institution_ids | https://openalex.org/I139759216 |
| authorships[1].affiliations[0].raw_affiliation_string | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| authorships[1].institutions[0].id | https://openalex.org/I139759216 |
| authorships[1].institutions[0].ror | https://ror.org/04w9fbh59 |
| authorships[1].institutions[0].type | education |
| authorships[1].institutions[0].lineage | https://openalex.org/I139759216 |
| authorships[1].institutions[0].country_code | CN |
| authorships[1].institutions[0].display_name | Beijing University of Posts and Telecommunications |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Hongxia Bie |
| authorships[1].is_corresponding | False |
| authorships[1].raw_affiliation_strings | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| authorships[2].author.id | https://openalex.org/A5070851446 |
| authorships[2].author.orcid | https://orcid.org/0000-0001-6408-2314 |
| authorships[2].author.display_name | J. Zhao |
| authorships[2].countries | CN |
| authorships[2].affiliations[0].institution_ids | https://openalex.org/I139759216 |
| authorships[2].affiliations[0].raw_affiliation_string | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| authorships[2].institutions[0].id | https://openalex.org/I139759216 |
| authorships[2].institutions[0].ror | https://ror.org/04w9fbh59 |
| authorships[2].institutions[0].type | education |
| authorships[2].institutions[0].lineage | https://openalex.org/I139759216 |
| authorships[2].institutions[0].country_code | CN |
| authorships[2].institutions[0].display_name | Beijing University of Posts and Telecommunications |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Zhao Jing |
| authorships[2].is_corresponding | False |
| authorships[2].raw_affiliation_strings | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| authorships[3].author.id | https://openalex.org/A5113186207 |
| authorships[3].author.orcid | |
| authorships[3].author.display_name | Yichen Zhi |
| authorships[3].countries | CN |
| authorships[3].affiliations[0].institution_ids | https://openalex.org/I139759216 |
| authorships[3].affiliations[0].raw_affiliation_string | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| authorships[3].institutions[0].id | https://openalex.org/I139759216 |
| authorships[3].institutions[0].ror | https://ror.org/04w9fbh59 |
| authorships[3].institutions[0].type | education |
| authorships[3].institutions[0].lineage | https://openalex.org/I139759216 |
| authorships[3].institutions[0].country_code | CN |
| authorships[3].institutions[0].display_name | Beijing University of Posts and Telecommunications |
| authorships[3].author_position | last |
| authorships[3].raw_author_name | Yichen Zhi |
| authorships[3].is_corresponding | False |
| authorships[3].raw_affiliation_strings | Intelligent Media Computing Center, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China |
| has_content.pdf | True |
| has_content.grobid_xml | True |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://www.mdpi.com/2673-2688/6/6/108/pdf?version=1747918810 |
| open_access.oa_status | gold |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-10T00:00:00 |
| display_name | Scrub-and-Learn: Category-Aware Weight Modification for Machine Unlearning |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T03:46:38.306776 |
| primary_topic.id | https://openalex.org/T12072 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.9217000007629395 |
| primary_topic.domain.id | https://openalex.org/domains/3 |
| primary_topic.domain.display_name | Physical Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/1702 |
| primary_topic.subfield.display_name | Artificial Intelligence |
| primary_topic.display_name | Machine Learning and Algorithms |
| related_works | https://openalex.org/W4391375266, https://openalex.org/W2899084033, https://openalex.org/W2748952813, https://openalex.org/W2390279801, https://openalex.org/W4391913857, https://openalex.org/W2358668433, https://openalex.org/W4396701345, https://openalex.org/W2376932109, https://openalex.org/W2001405890, https://openalex.org/W4396696052 |
| cited_by_count | 0 |
| locations_count | 2 |
| best_oa_location.id | doi:10.3390/ai6060108 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4210200511 |
| best_oa_location.source.issn | 2673-2688 |
| best_oa_location.source.type | journal |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | 2673-2688 |
| best_oa_location.source.is_core | True |
| best_oa_location.source.is_in_doaj | True |
| best_oa_location.source.display_name | AI |
| best_oa_location.source.host_organization | https://openalex.org/P4310310987 |
| best_oa_location.source.host_organization_name | Multidisciplinary Digital Publishing Institute |
| best_oa_location.source.host_organization_lineage | https://openalex.org/P4310310987 |
| best_oa_location.source.host_organization_lineage_names | Multidisciplinary Digital Publishing Institute |
| best_oa_location.license | cc-by |
| best_oa_location.pdf_url | https://www.mdpi.com/2673-2688/6/6/108/pdf?version=1747918810 |
| best_oa_location.version | publishedVersion |
| best_oa_location.raw_type | journal-article |
| best_oa_location.license_id | https://openalex.org/licenses/cc-by |
| best_oa_location.is_accepted | True |
| best_oa_location.is_published | True |
| best_oa_location.raw_source_name | AI |
| best_oa_location.landing_page_url | https://doi.org/10.3390/ai6060108 |
| primary_location.id | doi:10.3390/ai6060108 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4210200511 |
| primary_location.source.issn | 2673-2688 |
| primary_location.source.type | journal |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | 2673-2688 |
| primary_location.source.is_core | True |
| primary_location.source.is_in_doaj | True |
| primary_location.source.display_name | AI |
| primary_location.source.host_organization | https://openalex.org/P4310310987 |
| primary_location.source.host_organization_name | Multidisciplinary Digital Publishing Institute |
| primary_location.source.host_organization_lineage | https://openalex.org/P4310310987 |
| primary_location.source.host_organization_lineage_names | Multidisciplinary Digital Publishing Institute |
| primary_location.license | cc-by |
| primary_location.pdf_url | https://www.mdpi.com/2673-2688/6/6/108/pdf?version=1747918810 |
| primary_location.version | publishedVersion |
| primary_location.raw_type | journal-article |
| primary_location.license_id | https://openalex.org/licenses/cc-by |
| primary_location.is_accepted | True |
| primary_location.is_published | True |
| primary_location.raw_source_name | AI |
| primary_location.landing_page_url | https://doi.org/10.3390/ai6060108 |
| publication_date | 2025-05-22 |
| publication_year | 2025 |
| referenced_works | https://openalex.org/W4394595144, https://openalex.org/W4366371549, https://openalex.org/W4404293797, https://openalex.org/W4409563611, https://openalex.org/W4407360741, https://openalex.org/W4393213139, https://openalex.org/W4382559883, https://openalex.org/W4405660987, https://openalex.org/W6854470969, https://openalex.org/W4387986864, https://openalex.org/W4390571451, https://openalex.org/W6857734180, https://openalex.org/W6875877013, https://openalex.org/W4385806592, https://openalex.org/W3210087465, https://openalex.org/W4401078123, https://openalex.org/W4402175387, https://openalex.org/W6791907524, https://openalex.org/W6842523198, https://openalex.org/W4385955293, https://openalex.org/W4283323859, https://openalex.org/W6784265177, https://openalex.org/W6850390306, https://openalex.org/W4399801170, https://openalex.org/W4388505026, https://openalex.org/W4362714589, https://openalex.org/W3213758632, https://openalex.org/W3188298126, https://openalex.org/W3030364939, https://openalex.org/W4392173735, https://openalex.org/W2788388592, https://openalex.org/W3163939464, https://openalex.org/W4385064145, https://openalex.org/W2994887008, https://openalex.org/W3095058588, https://openalex.org/W2060277733, https://openalex.org/W2112796928, https://openalex.org/W6684191040, https://openalex.org/W2335728318, https://openalex.org/W2194775991, https://openalex.org/W3118608800, https://openalex.org/W2183341477, https://openalex.org/W3154155772 |
| referenced_works_count | 43 |
| abstract_inverted_index.a | 5, 60, 75, 123, 135 |
| abstract_inverted_index.4% | 136 |
| abstract_inverted_index.By | 71 |
| abstract_inverted_index.In | 52 |
| abstract_inverted_index.as | 34, 74 |
| abstract_inverted_index.in | 8, 15, 21 |
| abstract_inverted_index.is | 59 |
| abstract_inverted_index.of | 19, 84, 102, 131 |
| abstract_inverted_index.on | 112 |
| abstract_inverted_index.or | 39 |
| abstract_inverted_index.to | 42, 66, 90 |
| abstract_inverted_index.we | 55 |
| abstract_inverted_index.(1) | 0 |
| abstract_inverted_index.(2) | 50 |
| abstract_inverted_index.(3) | 106 |
| abstract_inverted_index.and | 11, 108 |
| abstract_inverted_index.our | 79, 117 |
| abstract_inverted_index.the | 36, 43, 87, 100, 103, 129, 139 |
| abstract_inverted_index.data | 20 |
| abstract_inverted_index.from | 86, 138 |
| abstract_inverted_index.high | 30 |
| abstract_inverted_index.rate | 125 |
| abstract_inverted_index.rest | 101 |
| abstract_inverted_index.role | 7 |
| abstract_inverted_index.such | 33 |
| abstract_inverted_index.that | 116 |
| abstract_inverted_index.this | 53 |
| abstract_inverted_index.with | 29 |
| abstract_inverted_index.below | 126 |
| abstract_inverted_index.data, | 46 |
| abstract_inverted_index.guide | 91 |
| abstract_inverted_index.model | 12 |
| abstract_inverted_index.often | 27 |
| abstract_inverted_index.plays | 4 |
| abstract_inverted_index.task, | 78 |
| abstract_inverted_index.their | 48 |
| abstract_inverted_index.which | 58 |
| abstract_inverted_index.while | 98 |
| abstract_inverted_index.work, | 54 |
| abstract_inverted_index.access | 41 |
| abstract_inverted_index.costs, | 32 |
| abstract_inverted_index.entire | 17 |
| abstract_inverted_index.labels | 83 |
| abstract_inverted_index.method | 80, 118 |
| abstract_inverted_index.model. | 141 |
| abstract_inverted_index.remove | 67 |
| abstract_inverted_index.target | 88 |
| abstract_inverted_index.tasks. | 23 |
| abstract_inverted_index.weight | 62, 92 |
| abstract_inverted_index.within | 134 |
| abstract_inverted_index.Machine | 2 |
| abstract_inverted_index.Results | 107 |
| abstract_inverted_index.classes | 133 |
| abstract_inverted_index.crucial | 6 |
| abstract_inverted_index.inverse | 37 |
| abstract_inverted_index.methods | 26 |
| abstract_inverted_index.privacy | 9 |
| abstract_inverted_index.require | 40 |
| abstract_inverted_index.results | 111 |
| abstract_inverted_index.samples | 85 |
| abstract_inverted_index.Hessian, | 38 |
| abstract_inverted_index.However, | 24 |
| abstract_inverted_index.Methods: | 51 |
| abstract_inverted_index.category | 89 |
| abstract_inverted_index.designed | 65 |
| abstract_inverted_index.existing | 25 |
| abstract_inverted_index.learning | 77 |
| abstract_inverted_index.limiting | 47 |
| abstract_inverted_index.modeling | 72 |
| abstract_inverted_index.multiple | 113 |
| abstract_inverted_index.original | 44, 140 |
| abstract_inverted_index.retained | 132 |
| abstract_inverted_index.struggle | 28 |
| abstract_inverted_index.targeted | 121 |
| abstract_inverted_index.training | 45 |
| abstract_inverted_index.unwanted | 96 |
| abstract_inverted_index.updates, | 93 |
| abstract_inverted_index.capacity. | 105 |
| abstract_inverted_index.continual | 76 |
| abstract_inverted_index.deviation | 137 |
| abstract_inverted_index.framework | 64 |
| abstract_inverted_index.introduce | 56 |
| abstract_inverted_index.knowledge | 69, 97 |
| abstract_inverted_index.leverages | 81 |
| abstract_inverted_index.model’s | 104 |
| abstract_inverted_index.scrubbing | 95 |
| abstract_inverted_index.5%—while | 127 |
| abstract_inverted_index.benchmarks | 114 |
| abstract_inverted_index.categories | 18 |
| abstract_inverted_index.eliminates | 120 |
| abstract_inverted_index.estimating | 35 |
| abstract_inverted_index.forgetting | 16 |
| abstract_inverted_index.preserving | 99, 128 |
| abstract_inverted_index.protection | 10 |
| abstract_inverted_index.re-encoded | 82 |
| abstract_inverted_index.unlearning | 3, 73 |
| abstract_inverted_index.Background: | 1 |
| abstract_inverted_index.class-level | 68 |
| abstract_inverted_index.demonstrate | 115 |
| abstract_inverted_index.effectively | 94, 119 |
| abstract_inverted_index.performance | 130 |
| abstract_inverted_index.recognition | 124 |
| abstract_inverted_index.Conclusions: | 109 |
| abstract_inverted_index.Experimental | 110 |
| abstract_inverted_index.efficiently. | 70 |
| abstract_inverted_index.modification | 63 |
| abstract_inverted_index.particularly | 14 |
| abstract_inverted_index.computational | 31 |
| abstract_inverted_index.optimization, | 13 |
| abstract_inverted_index.practicality. | 49 |
| abstract_inverted_index.category-aware | 61 |
| abstract_inverted_index.classification | 22 |
| abstract_inverted_index.Scrub-and-Learn, | 57 |
| abstract_inverted_index.categories—achieving | 122 |
| cited_by_percentile_year | |
| countries_distinct_count | 1 |
| institutions_distinct_count | 4 |
| citation_normalized_percentile.value | 0.0776525 |
| citation_normalized_percentile.is_in_top_1_percent | False |
| citation_normalized_percentile.is_in_top_10_percent | True |