{"id":"https://openalex.org/W4416034080","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.840","title":"Identifying Noise in Human-Created Datasets using Training Dynamics from Generative Models","display_name":"Identifying Noise in Human-Created Datasets using Training Dynamics from Generative Models","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416034080","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.840"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.840","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.840","pdf_url":"https://aclanthology.org/2025.findings-emnlp.840.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-emnlp.840.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5085409429","display_name":"Maeda F. Hanafi","orcid":"https://orcid.org/0009-0005-6663-2320"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Maeda Hanafi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077046363","display_name":"Ishan Jindal","orcid":"https://orcid.org/0000-0002-8854-4189"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ishan Jindal","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006509575","display_name":"Yannis Katsis","orcid":"https://orcid.org/0000-0002-1733-6227"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yannis Katsis","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5105463891","display_name":"Lucian Popa","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lucian Popa","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5076801334","display_name":"Huaiyu Zhu","orcid":"https://orcid.org/0000-0001-6918-4088"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huaiyu Zhu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.4136,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.85578845,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"15534","last_page":"15550"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.27649998664855957,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.27649998664855957,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12205","display_name":"Time Series Analysis and Forecasting","score":0.04639999940991402,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.03750000149011612,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.5514000058174133},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.5449000000953674},{"id":"https://openalex.org/keywords/dynamics","display_name":"Dynamics (music)","score":0.47749999165534973},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.4016000032424927},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.39959999918937683},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.3434999883174896}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6283000111579895},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5916000008583069},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.5514000058174133},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.5449000000953674},{"id":"https://openalex.org/C145912823","wikidata":"https://www.wikidata.org/wiki/Q113558","display_name":"Dynamics (music)","level":2,"score":0.47749999165534973},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4016000032424927},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.39959999918937683},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3675000071525574},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.3434999883174896},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.328900009393692},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2694000005722046},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.26820001006126404},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2678999900817871},{"id":"https://openalex.org/C100675267","wikidata":"https://www.wikidata.org/wiki/Q1371624","display_name":"Background noise","level":2,"score":0.25839999318122864}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-emnlp.840","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.840","pdf_url":"https://aclanthology.org/2025.findings-emnlp.840.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.840","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.840","pdf_url":"https://aclanthology.org/2025.findings-emnlp.840.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416034080.pdf","grobid_xml":"https://content.openalex.org/works/W4416034080.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Instruction":[0],"fine-tuning":[1],"enhances":[2],"the":[3,129],"alignment":[4],"of":[5,131],"autoregressive":[6,81],"language":[7,82],"models":[8,37,83],"(ArLMs)":[9],"with":[10],"human":[11],"intent":[12],"but":[13],"relies":[14],"on":[15],"large-scale":[16],"annotated":[17],"datasets":[18],"prone":[19],"to":[20,43,46,59,64,102],"label":[21],"and":[22,80,87,105,113],"text":[23],"noise.In":[24],"this":[25],"paper,":[26],"we":[27],"show":[28],"that":[29,118],"existing":[30],"noise":[31,91],"detection":[32],"techniques":[33],"designed":[34],"for":[35,124],"autoencoder":[36,79],"(AeLMs)":[38],"do":[39],"not":[40],"directly":[41],"generalize":[42],"ArLMs":[44,133],"due":[45],"differences":[47],"in":[48,134],"learning":[49],"dynamics.We":[50],"propose":[51],"TDRANKER,":[52],"a":[53,121],"novel":[54],"approach":[55],"leveraging":[56],"training":[57],"dynamics":[58],"rank":[60],"datapoints":[61],"from":[62],"easy-to-learn":[63],"hard-to-learn,":[65],"effectively":[66],"identifying":[67],"noisy":[68],"instances.Our":[69],"method":[70],"demonstrates":[71],"robustness":[72],"across":[73,88],"multiple":[74],"model":[75,114],"architectures":[76],"covering":[77],"both":[78],"(GPT-2,":[84],"BERT,":[85],"LaMini-Cerebras-256M)":[86],"various":[89],"dataset":[90],"levels,":[92],"achieving":[93],"at":[94],"least":[95],"2x":[96],"faster":[97],"denoising":[98],"than":[99],"previous":[100],"techniques.Applied":[101],"real-world":[103],"classification":[104],"generative":[106],"tasks,":[107],"TDRANKER":[108,119],"significantly":[109],"improves":[110],"data":[111],"quality":[112],"performance.These":[115],"findings":[116],"suggest":[117],"provides":[120],"scalable":[122],"solution":[123],"refining":[125],"instructiontuning":[126],"datasets,":[127],"enhancing":[128],"reliability":[130],"finetuned":[132],"practical":[135],"applications.":[136]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-11-08T00:00:00"}