{"id":"https://openalex.org/W4412888546","doi":"https://doi.org/10.18653/v1/2025.findings-acl.506","title":"BiasGuard: A Reasoning-Enhanced Bias Detection Tool for Large Language Models","display_name":"BiasGuard: A Reasoning-Enhanced Bias Detection Tool for Large Language Models","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888546","doi":"https://doi.org/10.18653/v1/2025.findings-acl.506"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.506","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.506","pdf_url":"https://aclanthology.org/2025.findings-acl.506.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: ACL 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-acl.506.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Zhiting Fan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhiting Fan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5117428676","display_name":"Ruizhe Chen","orcid":"https://orcid.org/0000-0003-2302-6775"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ruizhe Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5024343415","display_name":"Zuozhu Liu","orcid":"https://orcid.org/0000-0002-7816-502X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zuozhu Liu","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":true,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"9753","last_page":"9764"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9688000082969666,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9688000082969666,"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":0.923799991607666,"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/computer-science","display_name":"Computer science","score":0.7103754281997681},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.4522373080253601},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4444710612297058},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.34655168652534485}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7103754281997681},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.4522373080253601},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4444710612297058},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.34655168652534485}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.506","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.506","pdf_url":"https://aclanthology.org/2025.findings-acl.506.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: ACL 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-acl.506","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.506","pdf_url":"https://aclanthology.org/2025.findings-acl.506.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: ACL 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G4474290519","display_name":"\u57fa\u4e8e\u591a\u6a21\u6001\u53e3\u8154CBCT \u4e0e\u53e3\u626b\u5927\u6570\u636e\u7684\u4e34\u5e8a\u53ef\u9002\u7528\u51a0\u6839\u9aa8\u4e09\u7ef4\u878d\u5408\u53ca\u91cd\u6784\u7b97\u6cd5\u7814\u7a76","funder_award_id":"LZ23F020008","funder_id":"https://openalex.org/F4320338464","funder_display_name":"Natural Science Foundation of Zhejiang Province"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320322927","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884"},{"id":"https://openalex.org/F4320338464","display_name":"Natural Science Foundation of Zhejiang Province","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888546.pdf","grobid_xml":"https://content.openalex.org/works/W4412888546.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Identifying":[0],"bias":[1,46],"in":[2,12,27],"LLM-generated":[3],"content":[4],"is":[5,62],"a":[6,44,65],"crucial":[7],"prerequisite":[8],"for":[9,36,122],"ensuring":[10],"fairness":[11,17,37,56,79],"LLMs.Existing":[13],"methods,":[14],"such":[15],"as":[16],"classifiers":[18],"and":[19,31,53,92,108,119],"LLM-based":[20],"judges,":[21],"face":[22],"limitations":[23],"related":[24],"to":[25,58,74,88],"difficulties":[26],"understanding":[28],"underlying":[29],"intentions":[30],"the":[32,68,72,82,114,123],"lack":[33],"of":[34,116,125],"criteria":[35],"judgment.In":[38],"this":[39],"paper,":[40],"we":[41],"introduce":[42],"Bias-Guard,":[43],"novel":[45],"detection":[47],"tool":[48],"that":[49,101],"explicitly":[50,75],"analyzes":[51],"inputs":[52],"reasons":[54],"through":[55,64],"specifications":[57],"provide":[59,120],"accurate":[60],"judgments.BiasGuard":[61],"implemented":[63],"two-stage":[66],"approach:":[67],"first":[69],"stage":[70,84],"initializes":[71],"model":[73],"reason":[76],"based":[77],"on":[78],"specifications,":[80],"while":[81],"second":[83],"leverages":[85],"reinforcement":[86],"learning":[87],"enhance":[89],"its":[90],"reasoning":[91],"judgment":[93],"capabilities.Our":[94],"experiments,":[95],"conducted":[96],"across":[97],"five":[98],"datasets,":[99],"demonstrate":[100],"Bi-asGuard":[102],"outperforms":[103],"existing":[104],"tools,":[105],"improving":[106],"accuracy":[107],"reducing":[109],"over-fairness":[110],"misjudgments.We":[111],"also":[112],"highlight":[113],"importance":[115],"reasoning-enhanced":[117],"decision-making":[118],"evidence":[121],"effectiveness":[124],"our":[126],"twostage":[127],"optimization":[128],"pipeline.":[129]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}