The Implicit Regularization of Cosine Similarity in Large Language Model Alignment Article Swipe
Revista, Zen
,
IA, 10
·
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.5281/zenodo.17818972
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.5281/zenodo.17818972
Large Language Models (LLMs) have demonstrated remarkable capabilities across various natural language processing tasks, yet their practical deployment necessitates robust alignment with human values and intentions. Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO) are leading paradigms for this alignment, often relying on reward models or preference functions that implicitly or explicitly leverage semantic similarity. This paper investigates the hypothesis that the inherent properties of cosine similarity, frequently employed within these alignment frameworks, act as a form of implicit regularization. We propose that by normalizing vector magnitudes and focusing solely on directional consistency in high-dimensional embedding spaces, cosine similarity inherently discourages overfitting to spurious features, promotes a more generalized understanding of "preferred" responses, and enhances the stability of the alignment process. Through a conceptual framework and an outline of a comparative experimental design, we argue that this implicit regularization contributes to improved generalization, robustness, and semantic coherence in aligned LLMs, mitigating issues such as reward hacking and catastrophic forgetting. Our analysis suggests that understanding and intentionally leveraging such implicit regularization mechanisms can lead to more effective and controllable LLM alignment strategies.
Related Topics
Concepts
Computer science
Overfitting
Spurious relationship
Artificial intelligence
Regularization (linguistics)
Embedding
Trigonometric functions
Leverage (statistics)
Reinforcement learning
Language model
Machine learning
Natural language
Stability (learning theory)
Similarity (geometry)
Cosine similarity
Theoretical computer science
Reuse
Algorithm
Discrete cosine transform
Probabilistic logic
Natural language processing
Coherence (philosophical gambling strategy)
Semantics (computer science)
Consistency (knowledge bases)
Inference
Statistical model
Metadata
- Type
- article
- Landing Page
- https://doi.org/10.5281/zenodo.17818972
- OA Status
- green
- OpenAlex ID
- https://openalex.org/W7108720281
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W7108720281Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.5281/zenodo.17818972Digital Object Identifier
- Title
-
The Implicit Regularization of Cosine Similarity in Large Language Model AlignmentWork title
- Type
-
articleOpenAlex work type
- Publication year
-
2025Year of publication
- Publication date
-
2025-12-04Full publication date if available
- Authors
-
Revista, Zen, IA, 10List of authors in order
- Landing page
-
https://doi.org/10.5281/zenodo.17818972Publisher landing page
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://doi.org/10.5281/zenodo.17818972Direct OA link when available
- Concepts
-
Computer science, Overfitting, Spurious relationship, Artificial intelligence, Regularization (linguistics), Embedding, Trigonometric functions, Leverage (statistics), Reinforcement learning, Language model, Machine learning, Natural language, Stability (learning theory), Similarity (geometry), Cosine similarity, Theoretical computer science, Reuse, Algorithm, Discrete cosine transform, Probabilistic logic, Natural language processing, Coherence (philosophical gambling strategy), Semantics (computer science), Consistency (knowledge bases), Inference, Statistical modelTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
0Total citation count in OpenAlex
Full payload
| id | https://openalex.org/W7108720281 |
|---|---|
| doi | https://doi.org/10.5281/zenodo.17818972 |
| ids.doi | https://doi.org/10.5281/zenodo.17818972 |
| ids.openalex | https://openalex.org/W7108720281 |
| fwci | 0.0 |
| type | article |
| title | The Implicit Regularization of Cosine Similarity in Large Language Model Alignment |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T10028 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.45399174094200134 |
| 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 | Topic Modeling |
| topics[1].id | https://openalex.org/T11636 |
| topics[1].field.id | https://openalex.org/fields/27 |
| topics[1].field.display_name | Medicine |
| topics[1].score | 0.07150058448314667 |
| topics[1].domain.id | https://openalex.org/domains/4 |
| topics[1].domain.display_name | Health Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/2718 |
| topics[1].subfield.display_name | Health Informatics |
| topics[1].display_name | Artificial Intelligence in Healthcare and Education |
| topics[2].id | https://openalex.org/T12026 |
| topics[2].field.id | https://openalex.org/fields/17 |
| topics[2].field.display_name | Computer Science |
| topics[2].score | 0.059490811079740524 |
| topics[2].domain.id | https://openalex.org/domains/3 |
| topics[2].domain.display_name | Physical Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/1702 |
| topics[2].subfield.display_name | Artificial Intelligence |
| topics[2].display_name | Explainable Artificial Intelligence (XAI) |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C41008148 |
| concepts[0].level | 0 |
| concepts[0].score | 0.6804648637771606 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[0].display_name | Computer science |
| concepts[1].id | https://openalex.org/C22019652 |
| concepts[1].level | 3 |
| concepts[1].score | 0.6197466850280762 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q331309 |
| concepts[1].display_name | Overfitting |
| concepts[2].id | https://openalex.org/C97256817 |
| concepts[2].level | 2 |
| concepts[2].score | 0.6007974147796631 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q1462316 |
| concepts[2].display_name | Spurious relationship |
| concepts[3].id | https://openalex.org/C154945302 |
| concepts[3].level | 1 |
| concepts[3].score | 0.5614998936653137 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q11660 |
| concepts[3].display_name | Artificial intelligence |
| concepts[4].id | https://openalex.org/C2776135515 |
| concepts[4].level | 2 |
| concepts[4].score | 0.5527563691139221 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q17143721 |
| concepts[4].display_name | Regularization (linguistics) |
| concepts[5].id | https://openalex.org/C41608201 |
| concepts[5].level | 2 |
| concepts[5].score | 0.4715542495250702 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q980509 |
| concepts[5].display_name | Embedding |
| concepts[6].id | https://openalex.org/C178009071 |
| concepts[6].level | 2 |
| concepts[6].score | 0.41748249530792236 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q93344 |
| concepts[6].display_name | Trigonometric functions |
| concepts[7].id | https://openalex.org/C153083717 |
| concepts[7].level | 2 |
| concepts[7].score | 0.4052436351776123 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q6535263 |
| concepts[7].display_name | Leverage (statistics) |
| concepts[8].id | https://openalex.org/C97541855 |
| concepts[8].level | 2 |
| concepts[8].score | 0.3999088704586029 |
| concepts[8].wikidata | https://www.wikidata.org/wiki/Q830687 |
| concepts[8].display_name | Reinforcement learning |
| concepts[9].id | https://openalex.org/C137293760 |
| concepts[9].level | 2 |
| concepts[9].score | 0.38760992884635925 |
| concepts[9].wikidata | https://www.wikidata.org/wiki/Q3621696 |
| concepts[9].display_name | Language model |
| concepts[10].id | https://openalex.org/C119857082 |
| concepts[10].level | 1 |
| concepts[10].score | 0.3742230534553528 |
| concepts[10].wikidata | https://www.wikidata.org/wiki/Q2539 |
| concepts[10].display_name | Machine learning |
| concepts[11].id | https://openalex.org/C195324797 |
| concepts[11].level | 2 |
| concepts[11].score | 0.36444440484046936 |
| concepts[11].wikidata | https://www.wikidata.org/wiki/Q33742 |
| concepts[11].display_name | Natural language |
| concepts[12].id | https://openalex.org/C112972136 |
| concepts[12].level | 2 |
| concepts[12].score | 0.3254956305027008 |
| concepts[12].wikidata | https://www.wikidata.org/wiki/Q7595718 |
| concepts[12].display_name | Stability (learning theory) |
| concepts[13].id | https://openalex.org/C103278499 |
| concepts[13].level | 3 |
| concepts[13].score | 0.3224111795425415 |
| concepts[13].wikidata | https://www.wikidata.org/wiki/Q254465 |
| concepts[13].display_name | Similarity (geometry) |
| concepts[14].id | https://openalex.org/C2780762811 |
| concepts[14].level | 3 |
| concepts[14].score | 0.31936949491500854 |
| concepts[14].wikidata | https://www.wikidata.org/wiki/Q1784941 |
| concepts[14].display_name | Cosine similarity |
| concepts[15].id | https://openalex.org/C80444323 |
| concepts[15].level | 1 |
| concepts[15].score | 0.3153763711452484 |
| concepts[15].wikidata | https://www.wikidata.org/wiki/Q2878974 |
| concepts[15].display_name | Theoretical computer science |
| concepts[16].id | https://openalex.org/C206588197 |
| concepts[16].level | 2 |
| concepts[16].score | 0.3103969097137451 |
| concepts[16].wikidata | https://www.wikidata.org/wiki/Q846574 |
| concepts[16].display_name | Reuse |
| concepts[17].id | https://openalex.org/C11413529 |
| concepts[17].level | 1 |
| concepts[17].score | 0.3083249032497406 |
| concepts[17].wikidata | https://www.wikidata.org/wiki/Q8366 |
| concepts[17].display_name | Algorithm |
| concepts[18].id | https://openalex.org/C2221639 |
| concepts[18].level | 3 |
| concepts[18].score | 0.30183714628219604 |
| concepts[18].wikidata | https://www.wikidata.org/wiki/Q2877 |
| concepts[18].display_name | Discrete cosine transform |
| concepts[19].id | https://openalex.org/C49937458 |
| concepts[19].level | 2 |
| concepts[19].score | 0.29072535037994385 |
| concepts[19].wikidata | https://www.wikidata.org/wiki/Q2599292 |
| concepts[19].display_name | Probabilistic logic |
| concepts[20].id | https://openalex.org/C204321447 |
| concepts[20].level | 1 |
| concepts[20].score | 0.2891748547554016 |
| concepts[20].wikidata | https://www.wikidata.org/wiki/Q30642 |
| concepts[20].display_name | Natural language processing |
| concepts[21].id | https://openalex.org/C2781181686 |
| concepts[21].level | 2 |
| concepts[21].score | 0.28302276134490967 |
| concepts[21].wikidata | https://www.wikidata.org/wiki/Q4226068 |
| concepts[21].display_name | Coherence (philosophical gambling strategy) |
| concepts[22].id | https://openalex.org/C184337299 |
| concepts[22].level | 2 |
| concepts[22].score | 0.2734493017196655 |
| concepts[22].wikidata | https://www.wikidata.org/wiki/Q1437428 |
| concepts[22].display_name | Semantics (computer science) |
| concepts[23].id | https://openalex.org/C2776436953 |
| concepts[23].level | 2 |
| concepts[23].score | 0.26997092366218567 |
| concepts[23].wikidata | https://www.wikidata.org/wiki/Q5163215 |
| concepts[23].display_name | Consistency (knowledge bases) |
| concepts[24].id | https://openalex.org/C2776214188 |
| concepts[24].level | 2 |
| concepts[24].score | 0.25843140482902527 |
| concepts[24].wikidata | https://www.wikidata.org/wiki/Q408386 |
| concepts[24].display_name | Inference |
| concepts[25].id | https://openalex.org/C114289077 |
| concepts[25].level | 2 |
| concepts[25].score | 0.2583944499492645 |
| concepts[25].wikidata | https://www.wikidata.org/wiki/Q3284399 |
| concepts[25].display_name | Statistical model |
| keywords[0].id | https://openalex.org/keywords/overfitting |
| keywords[0].score | 0.6197466850280762 |
| keywords[0].display_name | Overfitting |
| keywords[1].id | https://openalex.org/keywords/spurious-relationship |
| keywords[1].score | 0.6007974147796631 |
| keywords[1].display_name | Spurious relationship |
| keywords[2].id | https://openalex.org/keywords/regularization |
| keywords[2].score | 0.5527563691139221 |
| keywords[2].display_name | Regularization (linguistics) |
| keywords[3].id | https://openalex.org/keywords/embedding |
| keywords[3].score | 0.4715542495250702 |
| keywords[3].display_name | Embedding |
| keywords[4].id | https://openalex.org/keywords/trigonometric-functions |
| keywords[4].score | 0.41748249530792236 |
| keywords[4].display_name | Trigonometric functions |
| keywords[5].id | https://openalex.org/keywords/leverage |
| keywords[5].score | 0.4052436351776123 |
| keywords[5].display_name | Leverage (statistics) |
| keywords[6].id | https://openalex.org/keywords/reinforcement-learning |
| keywords[6].score | 0.3999088704586029 |
| keywords[6].display_name | Reinforcement learning |
| keywords[7].id | https://openalex.org/keywords/language-model |
| keywords[7].score | 0.38760992884635925 |
| keywords[7].display_name | Language model |
| language | |
| locations[0].id | doi:10.5281/zenodo.17818972 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400562 |
| 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 | Zenodo (CERN European Organization for Nuclear Research) |
| locations[0].source.host_organization | https://openalex.org/I67311998 |
| locations[0].source.host_organization_name | European Organization for Nuclear Research |
| locations[0].source.host_organization_lineage | https://openalex.org/I67311998 |
| locations[0].license | cc-by |
| locations[0].pdf_url | |
| locations[0].version | |
| locations[0].raw_type | article-journal |
| locations[0].license_id | https://openalex.org/licenses/cc-by |
| locations[0].is_accepted | False |
| locations[0].is_published | |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | https://doi.org/10.5281/zenodo.17818972 |
| indexed_in | datacite |
| authorships[0].author.id | |
| authorships[0].author.orcid | |
| authorships[0].author.display_name | Revista, Zen |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Revista, Zen |
| authorships[0].is_corresponding | True |
| authorships[1].author.id | |
| authorships[1].author.orcid | |
| authorships[1].author.display_name | IA, 10 |
| authorships[1].author_position | last |
| authorships[1].raw_author_name | IA, 10 |
| authorships[1].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://doi.org/10.5281/zenodo.17818972 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-12-05T00:00:00 |
| display_name | The Implicit Regularization of Cosine Similarity in Large Language Model Alignment |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-12-05T23:25:22.460635 |
| primary_topic.id | https://openalex.org/T10028 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.45399174094200134 |
| 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 | Topic Modeling |
| cited_by_count | 0 |
| locations_count | 1 |
| best_oa_location.id | doi:10.5281/zenodo.17818972 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400562 |
| 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 | Zenodo (CERN European Organization for Nuclear Research) |
| best_oa_location.source.host_organization | https://openalex.org/I67311998 |
| best_oa_location.source.host_organization_name | European Organization for Nuclear Research |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I67311998 |
| best_oa_location.license | cc-by |
| best_oa_location.pdf_url | |
| best_oa_location.version | |
| best_oa_location.raw_type | article-journal |
| best_oa_location.license_id | https://openalex.org/licenses/cc-by |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | https://doi.org/10.5281/zenodo.17818972 |
| primary_location.id | doi:10.5281/zenodo.17818972 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400562 |
| 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 | Zenodo (CERN European Organization for Nuclear Research) |
| primary_location.source.host_organization | https://openalex.org/I67311998 |
| primary_location.source.host_organization_name | European Organization for Nuclear Research |
| primary_location.source.host_organization_lineage | https://openalex.org/I67311998 |
| primary_location.license | cc-by |
| primary_location.pdf_url | |
| primary_location.version | |
| primary_location.raw_type | article-journal |
| primary_location.license_id | https://openalex.org/licenses/cc-by |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | https://doi.org/10.5281/zenodo.17818972 |
| publication_date | 2025-12-04 |
| publication_year | 2025 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 78, 109, 125, 132 |
| abstract_inverted_index.We | 83 |
| abstract_inverted_index.an | 129 |
| abstract_inverted_index.as | 77, 156 |
| abstract_inverted_index.by | 86 |
| abstract_inverted_index.in | 96, 150 |
| abstract_inverted_index.of | 67, 80, 113, 120, 131 |
| abstract_inverted_index.on | 45, 93 |
| abstract_inverted_index.or | 48, 53 |
| abstract_inverted_index.to | 105, 143, 176 |
| abstract_inverted_index.we | 136 |
| abstract_inverted_index.LLM | 181 |
| abstract_inverted_index.Our | 162 |
| abstract_inverted_index.act | 76 |
| abstract_inverted_index.and | 24, 32, 90, 116, 128, 147, 159, 167, 179 |
| abstract_inverted_index.are | 37 |
| abstract_inverted_index.can | 174 |
| abstract_inverted_index.for | 40 |
| abstract_inverted_index.the | 61, 64, 118, 121 |
| abstract_inverted_index.yet | 14 |
| abstract_inverted_index.This | 58 |
| abstract_inverted_index.form | 79 |
| abstract_inverted_index.from | 28 |
| abstract_inverted_index.have | 4 |
| abstract_inverted_index.lead | 175 |
| abstract_inverted_index.more | 110, 177 |
| abstract_inverted_index.such | 155, 170 |
| abstract_inverted_index.that | 51, 63, 85, 138, 165 |
| abstract_inverted_index.this | 41, 139 |
| abstract_inverted_index.with | 21 |
| abstract_inverted_index.(DPO) | 36 |
| abstract_inverted_index.Human | 29 |
| abstract_inverted_index.LLMs, | 152 |
| abstract_inverted_index.Large | 0 |
| abstract_inverted_index.argue | 137 |
| abstract_inverted_index.human | 22 |
| abstract_inverted_index.often | 43 |
| abstract_inverted_index.paper | 59 |
| abstract_inverted_index.their | 15 |
| abstract_inverted_index.these | 73 |
| abstract_inverted_index.(LLMs) | 3 |
| abstract_inverted_index.(RLHF) | 31 |
| abstract_inverted_index.Direct | 33 |
| abstract_inverted_index.Models | 2 |
| abstract_inverted_index.across | 8 |
| abstract_inverted_index.cosine | 68, 100 |
| abstract_inverted_index.issues | 154 |
| abstract_inverted_index.models | 47 |
| abstract_inverted_index.reward | 46, 157 |
| abstract_inverted_index.robust | 19 |
| abstract_inverted_index.solely | 92 |
| abstract_inverted_index.tasks, | 13 |
| abstract_inverted_index.values | 23 |
| abstract_inverted_index.vector | 88 |
| abstract_inverted_index.within | 72 |
| abstract_inverted_index.Through | 124 |
| abstract_inverted_index.aligned | 151 |
| abstract_inverted_index.design, | 135 |
| abstract_inverted_index.hacking | 158 |
| abstract_inverted_index.leading | 38 |
| abstract_inverted_index.natural | 10 |
| abstract_inverted_index.outline | 130 |
| abstract_inverted_index.propose | 84 |
| abstract_inverted_index.relying | 44 |
| abstract_inverted_index.spaces, | 99 |
| abstract_inverted_index.various | 9 |
| abstract_inverted_index.Feedback | 30 |
| abstract_inverted_index.Language | 1 |
| abstract_inverted_index.Learning | 27 |
| abstract_inverted_index.analysis | 163 |
| abstract_inverted_index.employed | 71 |
| abstract_inverted_index.enhances | 117 |
| abstract_inverted_index.focusing | 91 |
| abstract_inverted_index.implicit | 81, 140, 171 |
| abstract_inverted_index.improved | 144 |
| abstract_inverted_index.inherent | 65 |
| abstract_inverted_index.language | 11 |
| abstract_inverted_index.leverage | 55 |
| abstract_inverted_index.process. | 123 |
| abstract_inverted_index.promotes | 108 |
| abstract_inverted_index.semantic | 56, 148 |
| abstract_inverted_index.spurious | 106 |
| abstract_inverted_index.suggests | 164 |
| abstract_inverted_index.alignment | 20, 74, 122, 182 |
| abstract_inverted_index.coherence | 149 |
| abstract_inverted_index.effective | 178 |
| abstract_inverted_index.embedding | 98 |
| abstract_inverted_index.features, | 107 |
| abstract_inverted_index.framework | 127 |
| abstract_inverted_index.functions | 50 |
| abstract_inverted_index.paradigms | 39 |
| abstract_inverted_index.practical | 16 |
| abstract_inverted_index.stability | 119 |
| abstract_inverted_index.Preference | 34 |
| abstract_inverted_index.alignment, | 42 |
| abstract_inverted_index.conceptual | 126 |
| abstract_inverted_index.deployment | 17 |
| abstract_inverted_index.explicitly | 54 |
| abstract_inverted_index.frequently | 70 |
| abstract_inverted_index.hypothesis | 62 |
| abstract_inverted_index.implicitly | 52 |
| abstract_inverted_index.inherently | 102 |
| abstract_inverted_index.leveraging | 169 |
| abstract_inverted_index.magnitudes | 89 |
| abstract_inverted_index.mechanisms | 173 |
| abstract_inverted_index.mitigating | 153 |
| abstract_inverted_index.preference | 49 |
| abstract_inverted_index.processing | 12 |
| abstract_inverted_index.properties | 66 |
| abstract_inverted_index.remarkable | 6 |
| abstract_inverted_index.responses, | 115 |
| abstract_inverted_index.similarity | 101 |
| abstract_inverted_index."preferred" | 114 |
| abstract_inverted_index.comparative | 133 |
| abstract_inverted_index.consistency | 95 |
| abstract_inverted_index.contributes | 142 |
| abstract_inverted_index.directional | 94 |
| abstract_inverted_index.discourages | 103 |
| abstract_inverted_index.forgetting. | 161 |
| abstract_inverted_index.frameworks, | 75 |
| abstract_inverted_index.generalized | 111 |
| abstract_inverted_index.intentions. | 25 |
| abstract_inverted_index.normalizing | 87 |
| abstract_inverted_index.overfitting | 104 |
| abstract_inverted_index.robustness, | 146 |
| abstract_inverted_index.similarity, | 69 |
| abstract_inverted_index.similarity. | 57 |
| abstract_inverted_index.strategies. | 183 |
| abstract_inverted_index.Optimization | 35 |
| abstract_inverted_index.capabilities | 7 |
| abstract_inverted_index.catastrophic | 160 |
| abstract_inverted_index.controllable | 180 |
| abstract_inverted_index.demonstrated | 5 |
| abstract_inverted_index.experimental | 134 |
| abstract_inverted_index.investigates | 60 |
| abstract_inverted_index.necessitates | 18 |
| abstract_inverted_index.Reinforcement | 26 |
| abstract_inverted_index.intentionally | 168 |
| abstract_inverted_index.understanding | 112, 166 |
| abstract_inverted_index.regularization | 141, 172 |
| abstract_inverted_index.generalization, | 145 |
| abstract_inverted_index.regularization. | 82 |
| abstract_inverted_index.high-dimensional | 97 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 2 |
| citation_normalized_percentile.value | 0.92145979 |
| citation_normalized_percentile.is_in_top_1_percent | False |
| citation_normalized_percentile.is_in_top_10_percent | True |