{"id":"https://openalex.org/W4415230519","doi":"https://doi.org/10.1609/aies.v8i2.36676","title":"When in Doubt, Cascade: Towards Building Efficient and Capable Guardrails","display_name":"When in Doubt, Cascade: Towards Building Efficient and Capable Guardrails","publication_year":2025,"publication_date":"2025-10-15","ids":{"openalex":"https://openalex.org/W4415230519","doi":"https://doi.org/10.1609/aies.v8i2.36676"},"language":"en","primary_location":{"id":"doi:10.1609/aies.v8i2.36676","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aies.v8i2.36676","pdf_url":"https://ojs.aaai.org/index.php/AIES/article/download/36676/38814","source":{"id":"https://openalex.org/S5407048695","display_name":"Proceedings of the AAAI/ACM Conference on AI Ethics and Society","issn_l":"3065-8365","issn":["3065-8365"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society","raw_type":"journal-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"bronze","oa_url":"https://ojs.aaai.org/index.php/AIES/article/download/36676/38814","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5064475546","display_name":"Manish Nagireddy","orcid":"https://orcid.org/0000-0001-9245-2546"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Manish Nagireddy","raw_affiliation_strings":["IBM Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM Research","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5001662671","display_name":"Inkit Padhi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Inkit Padhi","raw_affiliation_strings":["IBM Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM Research","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030936455","display_name":"Soumya K. Ghosh","orcid":"https://orcid.org/0000-0001-8359-581X"},"institutions":[{"id":"https://openalex.org/I4210123531","display_name":"Merck Biopharma Co., Ltd. (Japan)","ror":"https://ror.org/0292awh62","country_code":"JP","type":"company","lineage":["https://openalex.org/I164928964","https://openalex.org/I4210123531"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Soumya Ghosh","raw_affiliation_strings":["Merck"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Merck","institution_ids":["https://openalex.org/I4210123531"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5060534465","display_name":"Prasanna Sattigeri","orcid":"https://orcid.org/0000-0003-4435-0486"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Prasanna Sattigeri","raw_affiliation_strings":["IBM Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM Research","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":13.8322,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.9872225,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":"8","issue":"2","first_page":"1812","last_page":"1821"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T14445","display_name":"Transportation Safety and Impact Analysis","score":0.9939000010490417,"subfield":{"id":"https://openalex.org/subfields/2205","display_name":"Civil and Structural Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T14445","display_name":"Transportation Safety and Impact Analysis","score":0.9939000010490417,"subfield":{"id":"https://openalex.org/subfields/2205","display_name":"Civil and Structural Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/suite","display_name":"Suite","score":0.6577000021934509},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.6380000114440918},{"id":"https://openalex.org/keywords/variety","display_name":"Variety (cybernetics)","score":0.517799973487854},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.40689998865127563},{"id":"https://openalex.org/keywords/detector","display_name":"Detector","score":0.3506999909877777},{"id":"https://openalex.org/keywords/test-suite","display_name":"Test suite","score":0.32199999690055847}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7289999723434448},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.6577000021934509},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.6380000114440918},{"id":"https://openalex.org/C136197465","wikidata":"https://www.wikidata.org/wiki/Q1729295","display_name":"Variety (cybernetics)","level":2,"score":0.517799973487854},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.40689998865127563},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3833000063896179},{"id":"https://openalex.org/C94915269","wikidata":"https://www.wikidata.org/wiki/Q1834857","display_name":"Detector","level":2,"score":0.3506999909877777},{"id":"https://openalex.org/C151552104","wikidata":"https://www.wikidata.org/wiki/Q7705809","display_name":"Test suite","level":4,"score":0.32199999690055847},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.3077000081539154},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3059000074863434},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.30300000309944153},{"id":"https://openalex.org/C3018397939","wikidata":"https://www.wikidata.org/wiki/Q3644502","display_name":"Open source","level":3,"score":0.2800999879837036},{"id":"https://openalex.org/C2778348673","wikidata":"https://www.wikidata.org/wiki/Q739302","display_name":"Production (economics)","level":2,"score":0.2718000113964081},{"id":"https://openalex.org/C182306322","wikidata":"https://www.wikidata.org/wiki/Q1779371","display_name":"Order (exchange)","level":2,"score":0.2655999958515167}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1609/aies.v8i2.36676","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aies.v8i2.36676","pdf_url":"https://ojs.aaai.org/index.php/AIES/article/download/36676/38814","source":{"id":"https://openalex.org/S5407048695","display_name":"Proceedings of the AAAI/ACM Conference on AI Ethics and Society","issn_l":"3065-8365","issn":["3065-8365"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1609/aies.v8i2.36676","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aies.v8i2.36676","pdf_url":"https://ojs.aaai.org/index.php/AIES/article/download/36676/38814","source":{"id":"https://openalex.org/S5407048695","display_name":"Proceedings of the AAAI/ACM Conference on AI Ethics and Society","issn_l":"3065-8365","issn":["3065-8365"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4415230519.pdf","grobid_xml":"https://content.openalex.org/works/W4415230519.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2,40],"(LLMs)":[3],"have":[4],"convincing":[5],"performance":[6,123,138],"in":[7,72,145],"a":[8,49,59,87,125,140],"variety":[9],"of":[10,36,58,70,76,127,142,163],"downstream":[11],"tasks.":[12],"However,":[13],"these":[14],"systems":[15],"are":[16,166],"prone":[17],"to":[18,30,100,120],"generating":[19],"undesirable":[20],"outputs":[21],"such":[22,32],"as":[23,66],"harmful":[24],"and":[25,90,103,116,147,154,169],"biased":[26],"text.":[27],"In":[28],"order":[29],"remedy":[31],"generations,":[33],"the":[34,56,67,73,143],"development":[35],"guardrail":[37,156],"(or":[38],"detector)":[39],"has":[41],"gained":[42],"traction.":[43],"Motivated":[44],"by":[45],"findings":[46],"from":[47],"developing":[48,152],"detector":[50],"for":[51],"social":[52,78],"bias,":[53],"we":[54,64,85,109],"adopt":[55],"notion":[57],"use-mention":[60],"distinction":[61],"-":[62],"which":[63,96,165],"identified":[65],"primary":[68],"source":[69,129],"under-performance":[71],"preliminary":[74],"versions":[75],"our":[77,134],"bias":[79],"detector.":[80],"Armed":[81],"with":[82,139],"this":[83,107],"information,":[84],"describe":[86],"fully":[88],"extensible":[89],"reproducible":[91],"synthetic":[92],"data":[93],"generation":[94],"pipeline":[95],"leverages":[97],"taxonomy-driven":[98],"instructions":[99],"create":[101],"targeted":[102],"labeled":[104],"data.":[105],"Using":[106],"pipeline,":[108],"generate":[110],"over":[111],"300K":[112],"unique":[113],"contrastive":[114],"samples":[115],"provide":[117],"extensive":[118],"experiments":[119],"systematically":[121],"evaluate":[122],"on":[124],"suite":[126],"open":[128],"datasets.":[130],"We":[131],"show":[132],"that":[133],"method":[135],"achieves":[136],"competitive":[137],"fraction":[141],"cost":[144],"compute":[146],"offers":[148],"insight":[149],"into":[150],"iteratively":[151],"efficient":[153],"capable":[155],"models.":[157],"Warning:":[158],"This":[159],"paper":[160],"contains":[161],"examples":[162],"text":[164],"toxic,":[167],"biased,":[168],"potentially":[170],"harmful.":[171]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-16T00:00:00"}