{"id":"https://openalex.org/W7125786010","doi":"https://doi.org/10.48550/arxiv.2601.17329","title":"Conformal Feedback Alignment: Quantifying Answer-Level Reliability for Robust LLM Alignment","display_name":"Conformal Feedback Alignment: Quantifying Answer-Level Reliability for Robust LLM Alignment","publication_year":2026,"publication_date":"2026-01-24","ids":{"openalex":"https://openalex.org/W7125786010","doi":"https://doi.org/10.48550/arxiv.2601.17329"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2601.17329","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.17329","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.2601.17329","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5123924419","display_name":"Tiejin Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Tiejin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123906273","display_name":"Xiaoou Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Xiaoou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124026406","display_name":"Vishnu Nandam","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nandam, Vishnu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123895175","display_name":"Kuan-Ru Liou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liou, Kuan-Ru","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5123940264","display_name":"Hua Wei","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei, Hua","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":0.2214999943971634,"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":0.2214999943971634,"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.12600000202655792,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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":0.07680000364780426,"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/robustness","display_name":"Robustness (evolution)","score":0.771399974822998},{"id":"https://openalex.org/keywords/weighting","display_name":"Weighting","score":0.6944000124931335},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.6553000211715698},{"id":"https://openalex.org/keywords/conformal-map","display_name":"Conformal map","score":0.6173999905586243},{"id":"https://openalex.org/keywords/pairwise-comparison","display_name":"Pairwise comparison","score":0.593999981880188},{"id":"https://openalex.org/keywords/conformal-coating","display_name":"Conformal coating","score":0.3779999911785126}],"concepts":[{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.771399974822998},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.6944000124931335},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.6553000211715698},{"id":"https://openalex.org/C98214594","wikidata":"https://www.wikidata.org/wiki/Q850275","display_name":"Conformal map","level":2,"score":0.6173999905586243},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6096000075340271},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.593999981880188},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3831999897956848},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.37860000133514404},{"id":"https://openalex.org/C2777421745","wikidata":"https://www.wikidata.org/wiki/Q5160245","display_name":"Conformal coating","level":3,"score":0.3779999911785126},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35910001397132874},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.352400004863739},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3330000042915344},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.28290000557899475},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.2711000144481659},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.267300009727478},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.25029999017715454}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2601.17329","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.17329","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.2601.17329","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.17329","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":{"Preference-based":[0],"alignment":[1,96],"like":[2],"Reinforcement":[3],"Learning":[4],"from":[5,10],"Human":[6],"Feedback":[7,46],"(RLHF)":[8],"learns":[9],"pairwise":[11],"preferences,":[12,25],"yet":[13],"the":[14,32,35,41,56],"labels":[15],"are":[16,116],"often":[17],"noisy":[18],"and":[19,75,85,98,109],"inconsistent.":[20],"Existing":[21],"uncertainty-aware":[22],"approaches":[23],"weight":[24],"but":[26],"ignore":[27],"a":[28,49],"more":[29,111],"fundamental":[30],"factor:":[31],"reliability":[33,66],"of":[34,59],"\\emph{answers}":[36],"being":[37],"compared.":[38],"To":[39],"address":[40],"problem,":[42],"we":[43],"propose":[44],"Conformal":[45,60],"Alignment":[47],"(CFA),":[48],"framework":[50],"that":[51,93,102],"grounds":[52],"preference":[53],"weighting":[54,108],"in":[55],"statistical":[57],"guarantees":[58],"Prediction":[61],"(CP).":[62],"CFA":[63,94],"quantifies":[64],"answer-level":[65],"by":[67],"constructing":[68],"conformal":[69],"prediction":[70],"sets":[71],"with":[72],"controllable":[73],"coverage":[74],"aggregates":[76],"these":[77],"reliabilities":[78],"into":[79],"principled":[80],"weights":[81],"for":[82],"both":[83],"DPO-":[84],"PPO-style":[86],"training.":[87],"Experiments":[88],"across":[89],"different":[90],"datasets":[91],"show":[92],"improves":[95],"robustness":[97],"data":[99],"efficiency,":[100],"highlighting":[101],"modeling":[103],"\\emph{answer-side}":[104],"uncertainty":[105],"complements":[106],"preference-level":[107],"yields":[110],"robust,":[112],"data-efficient":[113],"alignment.":[114],"Codes":[115],"provided":[117],"here.":[118]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-01-28T00:00:00"}