{"id":"https://openalex.org/W4412888233","doi":"https://doi.org/10.18653/v1/2025.findings-acl.639","title":"SGDPO: Self-Guided Direct Preference Optimization for Language Model Alignment","display_name":"SGDPO: Self-Guided Direct Preference Optimization for Language Model Alignment","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888233","doi":"https://doi.org/10.18653/v1/2025.findings-acl.639"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.639","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.639","pdf_url":"https://aclanthology.org/2025.findings-acl.639.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.639.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5110106146","display_name":"Wenqiao Zhu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wenqiao Zhu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100689701","display_name":"Liu Ji","orcid":"https://orcid.org/0000-0002-5386-3550"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100444637","display_name":"Lulu Wang","orcid":"https://orcid.org/0000-0002-6796-2533"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lulu Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100341520","display_name":"Jun Wu","orcid":"https://orcid.org/0009-0002-0730-8860"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jun Wu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5074865219","display_name":"Yulun Zhang","orcid":"https://orcid.org/0000-0002-2288-5079"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yulun Zhang","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":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"12366","last_page":"12383"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9991999864578247,"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":0.9991999864578247,"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":0.993399977684021,"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/T12031","display_name":"Speech and dialogue systems","score":0.9768999814987183,"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":0.7066061496734619},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.6331344842910767},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.42823946475982666},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4107832908630371},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.38232338428497314},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.13596266508102417},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.06703472137451172}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7066061496734619},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.6331344842910767},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.42823946475982666},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4107832908630371},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.38232338428497314},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.13596266508102417},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.06703472137451172}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.639","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.639","pdf_url":"https://aclanthology.org/2025.findings-acl.639.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.639","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.639","pdf_url":"https://aclanthology.org/2025.findings-acl.639.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":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888233.pdf","grobid_xml":"https://content.openalex.org/works/W4412888233.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":{"Direct":[0,57],"Preference":[1,58],"Optimization":[2,59],"(DPO)":[3],"is":[4,35],"broadly":[5],"utilized":[6],"for":[7,78],"aligning":[8],"Large":[9],"Language":[10],"Models":[11],"(LLMs)":[12],"with":[13],"human":[14],"values":[15],"because":[16],"of":[17,29,40,84,94,130],"its":[18,20,100],"flexibility.Despite":[19],"effectiveness,":[21],"it":[22],"has":[23],"been":[24],"observed":[25],"that":[26],"the":[27,38,70,74,82,116,119,128],"capability":[28],"DPO":[30,41],"to":[31,68,135],"generate":[32],"human-preferred":[33],"response":[34],"limited":[36],"and":[37,86,98,110,122,126],"results":[39,114,121],"are":[42],"far":[43],"from":[44],"resilient.To":[45],"address":[46],"these":[47],"limitations,":[48],"in":[49],"this":[50],"paper":[51],"we":[52,103],"propose":[53],"a":[54,65,90],"novel":[55],"Self-Guided":[56],"algorithm,":[60],"i.e.,":[61],"SGDPO,":[62],"which":[63],"incorporates":[64],"pilot":[66],"term":[67],"steer":[69],"gradient":[71],"flow":[72],"during":[73],"optimization":[75],"process,":[76],"allowing":[77],"fine-grained":[79],"control":[80],"over":[81],"updates":[83],"chosen":[85],"rejected":[87],"rewards.We":[88],"provide":[89],"detailed":[91],"theoretical":[92,124],"analysis":[93,125],"our":[95,123,131],"proposed":[96,132],"method":[97],"elucidate":[99],"operational":[101],"mechanism.Furthermore,":[102],"conduct":[104],"comprehensive":[105],"experiments":[106],"on":[107],"various":[108],"models":[109],"benchmarks.The":[111],"extensive":[112],"experimental":[113],"demonstrate":[115],"consistency":[117],"between":[118],"empirical":[120],"confirm":[127],"effectiveness":[129],"approach":[133],"(up":[134],"9.19%":[136],"higher":[137],"score).":[138]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}