{"id":"https://openalex.org/W7161814335","doi":"https://doi.org/10.48550/arxiv.2605.18814","title":"How Faithful Is Trajectory-Based Data Attribution? Error Sources, Remedies, and Practical Guidelines","display_name":"How Faithful Is Trajectory-Based Data Attribution? Error Sources, Remedies, and Practical Guidelines","publication_year":2026,"publication_date":"2026-05-12","ids":{"openalex":"https://openalex.org/W7161814335","doi":"https://doi.org/10.48550/arxiv.2605.18814"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.18814","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18814","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.18814","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136516624","display_name":"Junwei Deng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Deng, Junwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030390478","display_name":"Pingbang Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Pingbang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136529217","display_name":"Suliang Jin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jin, Suliang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136521809","display_name":"Hao Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Hao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048878880","display_name":"Jiachen T. Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Jiachen T.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101412768","display_name":"Shichang Zhang","orcid":"https://orcid.org/0000-0003-0954-5018"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Shichang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136590031","display_name":"Jiaqi W. Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Jiaqi W.","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":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.11509999632835388,"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"}},"topics":[{"id":"https://openalex.org/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.11509999632835388,"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"}},{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.07689999788999557,"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"}},{"id":"https://openalex.org/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.05380000174045563,"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/trajectory","display_name":"Trajectory","score":0.49549999833106995},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.4350999891757965},{"id":"https://openalex.org/keywords/word-error-rate","display_name":"Word error rate","score":0.4302999973297119},{"id":"https://openalex.org/keywords/model-selection","display_name":"Model selection","score":0.4296000003814697},{"id":"https://openalex.org/keywords/error-detection-and-correction","display_name":"Error detection and correction","score":0.4180000126361847},{"id":"https://openalex.org/keywords/errors-in-variables-models","display_name":"Errors-in-variables models","score":0.3668999969959259},{"id":"https://openalex.org/keywords/proxy","display_name":"Proxy (statistics)","score":0.3619000017642975}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7468000054359436},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.5077999830245972},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.49549999833106995},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4564000070095062},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.4350999891757965},{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.4302999973297119},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.4296000003814697},{"id":"https://openalex.org/C103088060","wikidata":"https://www.wikidata.org/wiki/Q1062839","display_name":"Error detection and correction","level":2,"score":0.4180000126361847},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4108999967575073},{"id":"https://openalex.org/C179024874","wikidata":"https://www.wikidata.org/wiki/Q5395728","display_name":"Errors-in-variables models","level":2,"score":0.3668999969959259},{"id":"https://openalex.org/C2780148112","wikidata":"https://www.wikidata.org/wiki/Q1432581","display_name":"Proxy (statistics)","level":2,"score":0.3619000017642975},{"id":"https://openalex.org/C3018824978","wikidata":"https://www.wikidata.org/wiki/Q2894891","display_name":"Error analysis","level":2,"score":0.3262999951839447},{"id":"https://openalex.org/C19619285","wikidata":"https://www.wikidata.org/wiki/Q196372","display_name":"Observational error","level":2,"score":0.3124000132083893},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.30169999599456787},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3012000024318695},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2842000126838684},{"id":"https://openalex.org/C2781170535","wikidata":"https://www.wikidata.org/wiki/Q30587856","display_name":"Noisy data","level":2,"score":0.27869999408721924},{"id":"https://openalex.org/C28761237","wikidata":"https://www.wikidata.org/wiki/Q7805321","display_name":"Time horizon","level":2,"score":0.2696000039577484},{"id":"https://openalex.org/C2778484313","wikidata":"https://www.wikidata.org/wiki/Q1172540","display_name":"Data stream","level":2,"score":0.25519999861717224}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.18814","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18814","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.18814","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18814","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Trajectory-based":[0],"data":[1,26,28,68,216],"attribution":[2,113],"methods":[3,110],"estimate":[4],"the":[5,15,59,85,105,115,124,168,174,179,188,202,244,252,259],"influence":[6,154],"of":[7,38,42,63,117],"training":[8,16],"samples":[9],"on":[10],"model":[11,31],"predictions":[12],"by":[13,218],"unrolling":[14],"trajectory.":[17],"They":[18],"are":[19],"widely":[20],"used":[21],"in":[22,66,147],"applications":[23],"such":[24],"as":[25,104,185],"selection,":[27],"valuation,":[29],"and":[30,50,77,93,152,162,182,191,221,243],"diagnosis,":[32],"but":[33],"there":[34],"is":[35],"a":[36,193,225,235],"lack":[37],"comprehensive":[39],"error":[40,64,87,171,189,195],"analysis":[41,62],"these":[43,210,256],"methods,":[44],"raising":[45],"concerns":[46],"about":[47],"method":[48],"faithfulness":[49],"hindering":[51],"reliable":[52],"deployment.":[53],"In":[54],"this":[55,230],"work,":[56],"we":[57,100,166,208],"provide":[58],"first":[60],"systematic":[61],"sources":[65],"trajectory-based":[67],"attribution,":[69],"together":[70],"with":[71,123,234,251],"concrete":[72],"remedies":[73],"to":[74,133,144],"mitigate":[75],"them":[76],"practical":[78,213],"guidelines":[79,214],"for":[80,120,136,215,266],"downstream":[81],"use.":[82],"We":[83,95,130],"organize":[84],"total":[86],"into":[88,212,261],"three":[89,97],"categories,":[90],"config-level,":[91],"algorithm-level,":[92],"system-level.":[94],"make":[96],"contributions.":[98],"First,":[99],"identify":[101,178],"optimizer":[102,128],"mismatch":[103],"dominant":[106],"config-level":[107],"error:":[108],"existing":[109],"derive":[111,192],"their":[112],"under":[114,224],"assumption":[116],"SGD,":[118],"even":[119],"models":[121],"trained":[122],"modern":[125],"de":[126],"facto":[127],"AdamW.":[129],"propose":[131],"AdamW-influence":[132],"fully":[134],"account":[135],"AdamW's":[137],"optimization":[138],"dynamics,":[139],"yielding":[140],"improvements":[141],"from":[142,173],"10%":[143],"over":[145],"300%":[146],"Spearman":[148],"correlation":[149],"between":[150],"estimated":[151],"ground-truth":[153],"across":[155],"four":[156],"settings":[157],"spanning":[158],"MLP,":[159],"CNN,":[160],"GPT-2,":[161],"Llama":[163],"3.2-1B.":[164],"Second,":[165],"isolate":[167],"remaining":[169],"algorithm-level":[170],"arising":[172],"first-order":[175],"Taylor":[176],"approximation,":[177],"learning":[180,253],"rate":[181],"trajectory":[183,204],"length":[184],"factors":[186],"governing":[187],"magnitude,":[190],"closed-form":[194],"proxy":[196],"that":[197],"can":[198,247],"be":[199,248],"evaluated":[200],"along":[201],"original":[203],"without":[205],"retraining.":[206],"Third,":[207],"translate":[209],"insights":[211],"selection":[217,233,264],"unifying":[219],"offline":[220],"online":[222,232],"strategies":[223],"K-step":[226],"look-ahead":[227],"framework.":[228],"Under":[229],"framework,":[231],"short":[236],"horizon":[237,246],"often":[238],"matches":[239],"or":[240],"exceeds":[241],"offline,":[242],"optimal":[245],"tuned":[249],"jointly":[250],"rate.":[254],"Together,":[255],"results":[257],"turn":[258],"framework":[260],"an":[262],"actionable":[263],"recipe":[265],"practitioners.":[267]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-21T00:00:00"}