Fine Tuning without Catastrophic Forgetting via Selective Low Rank Adaptation Article Swipe
Reza Akbarian Bafghi
,
Carden Bagwell
,
Avinash Ravichandran
,
Ashish Shrivastava
,
Maziar Raissi
·
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2501.15377
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2501.15377
Adapting deep learning models to new domains often requires computationally intensive retraining and risks catastrophic forgetting. While fine-tuning enables domain-specific adaptation, it can reduce robustness to distribution shifts, impacting out-of-distribution (OOD) performance. Pre-trained zero-shot models like CLIP offer strong generalization but may suffer degraded robustness after fine-tuning. Building on Task Adaptive Parameter Sharing (TAPS), we propose a simple yet effective extension as a parameter-efficient fine-tuning (PEFT) method, using an indicator function to selectively activate Low-Rank Adaptation (LoRA) blocks. Our approach minimizes knowledge loss, retains its generalization strengths under domain shifts, and significantly reduces computational costs compared to traditional fine-tuning. We demonstrate that effective fine-tuning can be achieved with as few as 5\% of active blocks, substantially improving efficiency. Evaluations on pre-trained models such as CLIP and DINO-ViT demonstrate our method's broad applicability and effectiveness in maintaining performance and knowledge retention.
Related Topics
Concepts
Metadata
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2501.15377
- https://arxiv.org/pdf/2501.15377
- OA Status
- green
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4406880392
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4406880392Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2501.15377Digital Object Identifier
- Title
-
Fine Tuning without Catastrophic Forgetting via Selective Low Rank AdaptationWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2025Year of publication
- Publication date
-
2025-01-26Full publication date if available
- Authors
-
Reza Akbarian Bafghi, Carden Bagwell, Avinash Ravichandran, Ashish Shrivastava, Maziar RaissiList of authors in order
- Landing page
-
https://arxiv.org/abs/2501.15377Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2501.15377Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2501.15377Direct OA link when available
- Concepts
-
Forgetting, Adaptation (eye), Rank (graph theory), Computer science, Psychology, Mathematics, Cognitive psychology, Neuroscience, CombinatoricsTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
0Total citation count in OpenAlex
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4406880392 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2501.15377 |
| ids.doi | https://doi.org/10.48550/arxiv.2501.15377 |
| ids.openalex | https://openalex.org/W4406880392 |
| fwci | |
| type | preprint |
| title | Fine Tuning without Catastrophic Forgetting via Selective Low Rank Adaptation |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T10320 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.9767000079154968 |
| 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 | Neural Networks and Applications |
| topics[1].id | https://openalex.org/T11307 |
| topics[1].field.id | https://openalex.org/fields/17 |
| topics[1].field.display_name | Computer Science |
| topics[1].score | 0.9567000269889832 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/1702 |
| topics[1].subfield.display_name | Artificial Intelligence |
| topics[1].display_name | Domain Adaptation and Few-Shot Learning |
| topics[2].id | https://openalex.org/T10627 |
| topics[2].field.id | https://openalex.org/fields/17 |
| topics[2].field.display_name | Computer Science |
| topics[2].score | 0.9437000155448914 |
| topics[2].domain.id | https://openalex.org/domains/3 |
| topics[2].domain.display_name | Physical Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/1707 |
| topics[2].subfield.display_name | Computer Vision and Pattern Recognition |
| topics[2].display_name | Advanced Image and Video Retrieval Techniques |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C7149132 |
| concepts[0].level | 2 |
| concepts[0].score | 0.8325953483581543 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q1377840 |
| concepts[0].display_name | Forgetting |
| concepts[1].id | https://openalex.org/C139807058 |
| concepts[1].level | 2 |
| concepts[1].score | 0.6055737733840942 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q352374 |
| concepts[1].display_name | Adaptation (eye) |
| concepts[2].id | https://openalex.org/C164226766 |
| concepts[2].level | 2 |
| concepts[2].score | 0.5600345730781555 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q7293202 |
| concepts[2].display_name | Rank (graph theory) |
| concepts[3].id | https://openalex.org/C41008148 |
| concepts[3].level | 0 |
| concepts[3].score | 0.3947504162788391 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[3].display_name | Computer science |
| concepts[4].id | https://openalex.org/C15744967 |
| concepts[4].level | 0 |
| concepts[4].score | 0.291526198387146 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q9418 |
| concepts[4].display_name | Psychology |
| concepts[5].id | https://openalex.org/C33923547 |
| concepts[5].level | 0 |
| concepts[5].score | 0.262641042470932 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q395 |
| concepts[5].display_name | Mathematics |
| concepts[6].id | https://openalex.org/C180747234 |
| concepts[6].level | 1 |
| concepts[6].score | 0.2612963318824768 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q23373 |
| concepts[6].display_name | Cognitive psychology |
| concepts[7].id | https://openalex.org/C169760540 |
| concepts[7].level | 1 |
| concepts[7].score | 0.12058719992637634 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q207011 |
| concepts[7].display_name | Neuroscience |
| concepts[8].id | https://openalex.org/C114614502 |
| concepts[8].level | 1 |
| concepts[8].score | 0.10408282279968262 |
| concepts[8].wikidata | https://www.wikidata.org/wiki/Q76592 |
| concepts[8].display_name | Combinatorics |
| keywords[0].id | https://openalex.org/keywords/forgetting |
| keywords[0].score | 0.8325953483581543 |
| keywords[0].display_name | Forgetting |
| keywords[1].id | https://openalex.org/keywords/adaptation |
| keywords[1].score | 0.6055737733840942 |
| keywords[1].display_name | Adaptation (eye) |
| keywords[2].id | https://openalex.org/keywords/rank |
| keywords[2].score | 0.5600345730781555 |
| keywords[2].display_name | Rank (graph theory) |
| keywords[3].id | https://openalex.org/keywords/computer-science |
| keywords[3].score | 0.3947504162788391 |
| keywords[3].display_name | Computer science |
| keywords[4].id | https://openalex.org/keywords/psychology |
| keywords[4].score | 0.291526198387146 |
| keywords[4].display_name | Psychology |
| keywords[5].id | https://openalex.org/keywords/mathematics |
| keywords[5].score | 0.262641042470932 |
| keywords[5].display_name | Mathematics |
| keywords[6].id | https://openalex.org/keywords/cognitive-psychology |
| keywords[6].score | 0.2612963318824768 |
| keywords[6].display_name | Cognitive psychology |
| keywords[7].id | https://openalex.org/keywords/neuroscience |
| keywords[7].score | 0.12058719992637634 |
| keywords[7].display_name | Neuroscience |
| keywords[8].id | https://openalex.org/keywords/combinatorics |
| keywords[8].score | 0.10408282279968262 |
| keywords[8].display_name | Combinatorics |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2501.15377 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400194 |
| locations[0].source.issn | |
| locations[0].source.type | repository |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | arXiv (Cornell University) |
| locations[0].source.host_organization | https://openalex.org/I205783295 |
| locations[0].source.host_organization_name | Cornell University |
| locations[0].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[0].license | |
| locations[0].pdf_url | https://arxiv.org/pdf/2501.15377 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2501.15377 |
| locations[1].id | doi:10.48550/arxiv.2501.15377 |
| locations[1].is_oa | True |
| locations[1].source.id | https://openalex.org/S4306400194 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | True |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | arXiv (Cornell University) |
| locations[1].source.host_organization | https://openalex.org/I205783295 |
| locations[1].source.host_organization_name | Cornell University |
| locations[1].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[1].license | |
| locations[1].pdf_url | |
| locations[1].version | |
| locations[1].raw_type | article |
| locations[1].license_id | |
| locations[1].is_accepted | False |
| locations[1].is_published | |
| locations[1].raw_source_name | |
| locations[1].landing_page_url | https://doi.org/10.48550/arxiv.2501.15377 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5092675602 |
| authorships[0].author.orcid | |
| authorships[0].author.display_name | Reza Akbarian Bafghi |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Bafghi, Reza Akbarian |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5116054015 |
| authorships[1].author.orcid | |
| authorships[1].author.display_name | Carden Bagwell |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Bagwell, Carden |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5113407629 |
| authorships[2].author.orcid | |
| authorships[2].author.display_name | Avinash Ravichandran |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Ravichandran, Avinash |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5002740493 |
| authorships[3].author.orcid | https://orcid.org/0000-0002-2465-0708 |
| authorships[3].author.display_name | Ashish Shrivastava |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Shrivastava, Ashish |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5012536010 |
| authorships[4].author.orcid | https://orcid.org/0000-0002-8467-4568 |
| authorships[4].author.display_name | Maziar Raissi |
| authorships[4].author_position | last |
| authorships[4].raw_author_name | Raissi, Maziar |
| authorships[4].is_corresponding | False |
| has_content.pdf | False |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2501.15377 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-10T00:00:00 |
| display_name | Fine Tuning without Catastrophic Forgetting via Selective Low Rank Adaptation |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T10320 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.9767000079154968 |
| 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 | Neural Networks and Applications |
| related_works | https://openalex.org/W4391375266, https://openalex.org/W2899084033, https://openalex.org/W2748952813, https://openalex.org/W4289718052, https://openalex.org/W2164121020, https://openalex.org/W2145559838, https://openalex.org/W2905319430, https://openalex.org/W3116498279, https://openalex.org/W4287549553, https://openalex.org/W4310285384 |
| cited_by_count | 0 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2501.15377 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400194 |
| best_oa_location.source.issn | |
| best_oa_location.source.type | repository |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | arXiv (Cornell University) |
| best_oa_location.source.host_organization | https://openalex.org/I205783295 |
| best_oa_location.source.host_organization_name | Cornell University |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| best_oa_location.license | |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2501.15377 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | http://arxiv.org/abs/2501.15377 |
| primary_location.id | pmh:oai:arXiv.org:2501.15377 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400194 |
| primary_location.source.issn | |
| primary_location.source.type | repository |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | arXiv (Cornell University) |
| primary_location.source.host_organization | https://openalex.org/I205783295 |
| primary_location.source.host_organization_name | Cornell University |
| primary_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| primary_location.license | |
| primary_location.pdf_url | https://arxiv.org/pdf/2501.15377 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2501.15377 |
| publication_date | 2025-01-26 |
| publication_year | 2025 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 56, 62 |
| abstract_inverted_index.We | 99 |
| abstract_inverted_index.an | 68 |
| abstract_inverted_index.as | 61, 108, 110, 123 |
| abstract_inverted_index.be | 105 |
| abstract_inverted_index.in | 134 |
| abstract_inverted_index.it | 21 |
| abstract_inverted_index.of | 112 |
| abstract_inverted_index.on | 48, 119 |
| abstract_inverted_index.to | 4, 25, 71, 96 |
| abstract_inverted_index.we | 54 |
| abstract_inverted_index.5\% | 111 |
| abstract_inverted_index.Our | 78 |
| abstract_inverted_index.and | 12, 90, 125, 132, 137 |
| abstract_inverted_index.but | 40 |
| abstract_inverted_index.can | 22, 104 |
| abstract_inverted_index.few | 109 |
| abstract_inverted_index.its | 84 |
| abstract_inverted_index.may | 41 |
| abstract_inverted_index.new | 5 |
| abstract_inverted_index.our | 128 |
| abstract_inverted_index.yet | 58 |
| abstract_inverted_index.CLIP | 36, 124 |
| abstract_inverted_index.Task | 49 |
| abstract_inverted_index.deep | 1 |
| abstract_inverted_index.like | 35 |
| abstract_inverted_index.such | 122 |
| abstract_inverted_index.that | 101 |
| abstract_inverted_index.with | 107 |
| abstract_inverted_index.(OOD) | 30 |
| abstract_inverted_index.While | 16 |
| abstract_inverted_index.after | 45 |
| abstract_inverted_index.broad | 130 |
| abstract_inverted_index.costs | 94 |
| abstract_inverted_index.loss, | 82 |
| abstract_inverted_index.offer | 37 |
| abstract_inverted_index.often | 7 |
| abstract_inverted_index.risks | 13 |
| abstract_inverted_index.under | 87 |
| abstract_inverted_index.using | 67 |
| abstract_inverted_index.(LoRA) | 76 |
| abstract_inverted_index.(PEFT) | 65 |
| abstract_inverted_index.active | 113 |
| abstract_inverted_index.domain | 88 |
| abstract_inverted_index.models | 3, 34, 121 |
| abstract_inverted_index.reduce | 23 |
| abstract_inverted_index.simple | 57 |
| abstract_inverted_index.strong | 38 |
| abstract_inverted_index.suffer | 42 |
| abstract_inverted_index.(TAPS), | 53 |
| abstract_inverted_index.Sharing | 52 |
| abstract_inverted_index.blocks, | 114 |
| abstract_inverted_index.blocks. | 77 |
| abstract_inverted_index.domains | 6 |
| abstract_inverted_index.enables | 18 |
| abstract_inverted_index.method, | 66 |
| abstract_inverted_index.propose | 55 |
| abstract_inverted_index.reduces | 92 |
| abstract_inverted_index.retains | 83 |
| abstract_inverted_index.shifts, | 27, 89 |
| abstract_inverted_index.Adapting | 0 |
| abstract_inverted_index.Adaptive | 50 |
| abstract_inverted_index.Building | 47 |
| abstract_inverted_index.DINO-ViT | 126 |
| abstract_inverted_index.Low-Rank | 74 |
| abstract_inverted_index.achieved | 106 |
| abstract_inverted_index.activate | 73 |
| abstract_inverted_index.approach | 79 |
| abstract_inverted_index.compared | 95 |
| abstract_inverted_index.degraded | 43 |
| abstract_inverted_index.function | 70 |
| abstract_inverted_index.learning | 2 |
| abstract_inverted_index.method's | 129 |
| abstract_inverted_index.requires | 8 |
| abstract_inverted_index.Parameter | 51 |
| abstract_inverted_index.effective | 59, 102 |
| abstract_inverted_index.extension | 60 |
| abstract_inverted_index.impacting | 28 |
| abstract_inverted_index.improving | 116 |
| abstract_inverted_index.indicator | 69 |
| abstract_inverted_index.intensive | 10 |
| abstract_inverted_index.knowledge | 81, 138 |
| abstract_inverted_index.minimizes | 80 |
| abstract_inverted_index.strengths | 86 |
| abstract_inverted_index.zero-shot | 33 |
| abstract_inverted_index.Adaptation | 75 |
| abstract_inverted_index.retention. | 139 |
| abstract_inverted_index.retraining | 11 |
| abstract_inverted_index.robustness | 24, 44 |
| abstract_inverted_index.Evaluations | 118 |
| abstract_inverted_index.Pre-trained | 32 |
| abstract_inverted_index.adaptation, | 20 |
| abstract_inverted_index.demonstrate | 100, 127 |
| abstract_inverted_index.efficiency. | 117 |
| abstract_inverted_index.fine-tuning | 17, 64, 103 |
| abstract_inverted_index.forgetting. | 15 |
| abstract_inverted_index.maintaining | 135 |
| abstract_inverted_index.performance | 136 |
| abstract_inverted_index.pre-trained | 120 |
| abstract_inverted_index.selectively | 72 |
| abstract_inverted_index.traditional | 97 |
| abstract_inverted_index.catastrophic | 14 |
| abstract_inverted_index.distribution | 26 |
| abstract_inverted_index.fine-tuning. | 46, 98 |
| abstract_inverted_index.performance. | 31 |
| abstract_inverted_index.applicability | 131 |
| abstract_inverted_index.computational | 93 |
| abstract_inverted_index.effectiveness | 133 |
| abstract_inverted_index.significantly | 91 |
| abstract_inverted_index.substantially | 115 |
| abstract_inverted_index.generalization | 39, 85 |
| abstract_inverted_index.computationally | 9 |
| abstract_inverted_index.domain-specific | 19 |
| abstract_inverted_index.out-of-distribution | 29 |
| abstract_inverted_index.parameter-efficient | 63 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 5 |
| citation_normalized_percentile |