{"id":"https://openalex.org/W7161057253","doi":"https://doi.org/10.48550/arxiv.2605.11632","title":"Macro: Enhancing Multilingual Counterfactual Explanations through Alignment-as-Preference Optimization","display_name":"Macro: Enhancing Multilingual Counterfactual Explanations through Alignment-as-Preference Optimization","publication_year":2026,"publication_date":"2026-05-12","ids":{"openalex":"https://openalex.org/W7161057253","doi":"https://doi.org/10.48550/arxiv.2605.11632"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.11632","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.11632","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":null,"license_id":null,"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.2605.11632","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136048016","display_name":"Yilong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yilong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136003771","display_name":"Qianli Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Qianli","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122330490","display_name":"Bohao Chu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chu, Bohao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136048080","display_name":"Yihong Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yihong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136033707","display_name":"Jing Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Jing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136008162","display_name":"Simon Ostermann","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ostermann, Simon","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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.6617000102996826,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.6617000102996826,"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/T10028","display_name":"Topic Modeling","score":0.05040000006556511,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.038100000470876694,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/counterfactual-thinking","display_name":"Counterfactual thinking","score":0.8233000040054321},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.7645999789237976},{"id":"https://openalex.org/keywords/macro","display_name":"Macro","score":0.7105000019073486},{"id":"https://openalex.org/keywords/preference-elicitation","display_name":"Preference elicitation","score":0.40869998931884766},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.3528999984264374},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.35030001401901245}],"concepts":[{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.8233000040054321},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.7645999789237976},{"id":"https://openalex.org/C166955791","wikidata":"https://www.wikidata.org/wiki/Q629579","display_name":"Macro","level":2,"score":0.7105000019073486},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5340999960899353},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47440001368522644},{"id":"https://openalex.org/C2777868144","wikidata":"https://www.wikidata.org/wiki/Q7239817","display_name":"Preference elicitation","level":3,"score":0.40869998931884766},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3824000060558319},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.37119999527931213},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3601999878883362},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.3528999984264374},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.35030001401901245},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.3248000144958496},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.31049999594688416},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.3005000054836273},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.29679998755455017},{"id":"https://openalex.org/C181204326","wikidata":"https://www.wikidata.org/wiki/Q7239820","display_name":"Preference learning","level":3,"score":0.28209999203681946},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2635999917984009},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.2599000036716461}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.11632","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.11632","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.11632","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.11632","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.5533644556999207}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Self-generated":[0],"counterfactual":[1],"explanations":[2],"(SCEs)":[3],"are":[4],"minimally":[5],"modified":[6],"inputs":[7],"(minimality)":[8],"generated":[9],"by":[10,108],"large":[11],"language":[12],"models":[13],"(LLMs)":[14],"that":[15,66,85,104,141,154],"flip":[16],"their":[17],"own":[18],"predictions":[19],"(validity),":[20],"offering":[21],"a":[22,49,62,77,171],"causally":[23],"grounded":[24],"approach":[25],"to":[26,41,72,81,130],"unraveling":[27],"black-box":[28],"LLM":[29],"behavior.":[30],"Yet":[31],"extending":[32],"them":[33],"beyond":[34],"English":[35],"remains":[36],"challenging:":[37],"existing":[38],"methods":[39],"struggle":[40],"produce":[42],"valid":[43],"SCEs":[44],"in":[45],"non-dominant":[46],"languages,":[47],"and":[48,54,98,160],"persistent":[50],"trade-off":[51,89],"between":[52],"validity":[53,107],"minimality":[55,123],"undermines":[56],"explanation":[57],"quality.":[58],"We":[59],"introduce":[60],"Macro,":[61],"preference":[63,83,92,143,168],"alignment":[64,159],"framework":[65],"applies":[67],"Direct":[68],"Preference":[69],"Optimization":[70],"(DPO)":[71],"multilingual":[73,176],"SCE":[74],"generation,":[75],"using":[76],"composite":[78],"scoring":[79],"function":[80],"construct":[82],"pairs":[84],"effectively":[86],"translate":[87],"the":[88,113,121,126],"into":[90],"measurable":[91],"signals.":[93],"Experiments":[94],"across":[95],"four":[96],"LLMs":[97],"seven":[99],"typologically":[100],"diverse":[101],"languages":[102],"show":[103],"Macro":[105,133,155],"improves":[106],"12.55\\%":[109],"on":[110,137],"average":[111],"over":[112],"chain-of-thought":[114],"baseline":[115],"without":[116],"degrading":[117],"minimality,":[118],"while":[119],"avoiding":[120],"severe":[122],"violations":[124],"of":[125],"translation-based":[127],"baseline.":[128],"Compared":[129],"supervised":[131],"fine-tuning,":[132],"achieves":[134],"superior":[135],"performance":[136],"both":[138],"metrics,":[139],"confirming":[140],"explicit":[142],"optimization":[144,169],"is":[145],"essential":[146],"for":[147,174],"balancing":[148],"this":[149],"trade-off.":[150],"Further":[151],"analyses":[152],"reveal":[153],"increases":[156],"cross-lingual":[157],"perturbation":[158],"mitigates":[161],"common":[162],"generation":[163],"errors.":[164],"Our":[165],"results":[166],"highlight":[167],"as":[170],"promising":[172],"direction":[173],"enhancing":[175],"model":[177],"explanations.":[178]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-14T00:00:00"}