{"id":"https://openalex.org/W7124676945","doi":"https://doi.org/10.48550/arxiv.2601.11021","title":"Combating Spurious Correlations in Graph Interpretability via Self-Reflection","display_name":"Combating Spurious Correlations in Graph Interpretability via Self-Reflection","publication_year":2026,"publication_date":"2026-01-16","ids":{"openalex":"https://openalex.org/W7124676945","doi":"https://doi.org/10.48550/arxiv.2601.11021"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2601.11021","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.11021","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.2601.11021","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5123342475","display_name":"Kecheng Cai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cai, Kecheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123350256","display_name":"Chenyang Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Chenyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Peng, Chao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Peng, Chao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Huang, Jiafu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Jiafu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123827390","display_name":"Qiyuan Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liang, Qiyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5098550523","display_name":"Irene Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Irene","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":0.6728000044822693,"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":0.6728000044822693,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.25760000944137573,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.017000000923871994,"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/interpretability","display_name":"Interpretability","score":0.9733999967575073},{"id":"https://openalex.org/keywords/spurious-relationship","display_name":"Spurious relationship","score":0.8851000070571899},{"id":"https://openalex.org/keywords/graph","display_name":"Graph","score":0.6173999905586243},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.588699996471405},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.4309999942779541},{"id":"https://openalex.org/keywords/focus","display_name":"Focus (optics)","score":0.41100001335144043},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.4000999927520752}],"concepts":[{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.9733999967575073},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.8851000070571899},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7059000134468079},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.6173999905586243},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6122000217437744},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.588699996471405},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.580299973487854},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4309999942779541},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.41100001335144043},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.4000999927520752},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.36910000443458557},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.3614000082015991},{"id":"https://openalex.org/C62611344","wikidata":"https://www.wikidata.org/wiki/Q1062658","display_name":"Node (physics)","level":2,"score":0.35030001401901245},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.30559998750686646},{"id":"https://openalex.org/C106937863","wikidata":"https://www.wikidata.org/wiki/Q7236518","display_name":"Power graph analysis","level":3,"score":0.2994000017642975},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.29179999232292175},{"id":"https://openalex.org/C143587482","wikidata":"https://www.wikidata.org/wiki/Q1543216","display_name":"Iterative and incremental development","level":2,"score":0.2872999906539917},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.25600001215934753}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2601.11021","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.11021","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.2601.11021","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.11021","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":{"Interpretable":[0],"graph":[1,27,35,171,231],"learning":[2,172],"has":[3],"recently":[4],"emerged":[5],"as":[6],"a":[7,33,101,161,176,198,239],"popular":[8],"research":[9],"topic":[10],"in":[11,45,75,136,152],"machine":[12],"learning.":[13],"The":[14,73],"goal":[15],"is":[16,65],"to":[17,55,82,92,113,140,149,196,214,237],"identify":[18],"the":[19,62,66,124,131,193,222,228],"important":[20],"nodes":[21],"and":[22,48,184],"edges":[23],"of":[24,40,61,201,230],"an":[25],"input":[26],"that":[28,130,164],"are":[29,79],"crucial":[30],"for":[31,90,181],"performing":[32],"specific":[34],"reasoning":[36],"task.":[37],"A":[38],"number":[39],"studies":[41],"have":[42,52],"been":[43,53],"conducted":[44],"this":[46,76,110,116,225,245],"area,":[47],"various":[49],"benchmark":[50,78,111],"datasets":[51,74,153],"proposed":[54],"facilitate":[56],"evaluation.":[57,202],"Among":[58],"them,":[59],"one":[60],"most":[63],"challenging":[64,125],"Spurious-Motif":[67,126],"benchmark,":[68],"introduced":[69],"at":[70],"ICLR":[71],"2022.":[72],"synthetic":[77],"deliberately":[80],"designed":[81],"include":[83],"spurious":[84,156],"correlations,":[85],"making":[86],"it":[87],"particularly":[88],"difficult":[89],"models":[91,139,210],"distinguish":[93],"truly":[94],"relevant":[95],"structures":[96],"from":[97,227],"misleading":[98],"patterns.":[99],"As":[100],"result,":[102],"existing":[103,169],"methods":[104],"exhibit":[105],"significantly":[106],"worse":[107],"performance":[108],"on":[109,120,123,244],"compared":[112],"others.":[114],"In":[115],"paper,":[117],"we":[118,159],"focus":[119],"improving":[121],"interpretability":[122,151],"datasets.":[127],"We":[128,219],"demonstrate":[129],"self-reflection":[132,162],"technique,":[133],"commonly":[134],"used":[135],"large":[137,208],"language":[138,209],"tackle":[141],"complex":[142],"tasks,":[143],"can":[144,165],"also":[145],"be":[146,166],"effectively":[147],"adapted":[148],"enhance":[150],"with":[154,168],"strong":[155],"correlations.":[157],"Specifically,":[158],"propose":[160,238],"framework":[163,187],"integrated":[167],"interpretable":[170],"methods.":[173],"When":[174],"such":[175],"method":[177,195,242],"produces":[178],"importance":[179],"scores":[180],"each":[182],"node":[183],"edge,":[185],"our":[186],"feeds":[188],"these":[189],"predictions":[190],"back":[191],"into":[192],"original":[194],"perform":[197],"second":[199],"round":[200],"This":[203],"iterative":[204],"process":[205],"mirrors":[206],"how":[207],"employ":[211],"self-reflective":[212],"prompting":[213],"reassess":[215],"their":[216],"previous":[217],"outputs.":[218],"further":[220],"analyze":[221],"reasons":[223],"behind":[224],"improvement":[226],"perspective":[229],"representation":[232],"learning,":[233],"which":[234],"motivates":[235],"us":[236],"fine-tuning":[240],"training":[241],"based":[243],"feedback":[246],"mechanism.":[247]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-01-20T00:00:00"}