{"id":"https://openalex.org/W7225148442","doi":"https://doi.org/10.48550/arxiv.2611.02667","title":"UPSHIFT: User Preference Steering from Human-AI Interaction Feedback Traces","display_name":"UPSHIFT: User Preference Steering from Human-AI Interaction Feedback Traces","publication_year":2026,"publication_date":"2026-11-02","ids":{"openalex":"https://openalex.org/W7225148442","doi":"https://doi.org/10.48550/arxiv.2611.02667"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2611.02667","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2611.02667","pdf_url":"https://arxiv.org/pdf/2611.02667","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2611.02667","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5159606632","display_name":"Jingping Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]},{"id":"https://openalex.org/I4391012619","display_name":"Shanghai Artificial Intelligence Laboratory","ror":"https://ror.org/03wkvpx79","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4391012619"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wang, Jingping","raw_affiliation_strings":["Shanghai Artificial Intelligence Laboratory (shlab.org.cn","Shanghai Jiaotong University (sjtu"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Artificial Intelligence Laboratory (shlab.org.cn","institution_ids":["https://openalex.org/I4391012619"]},{"raw_affiliation_string":"Shanghai Jiaotong University (sjtu","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5149230687","display_name":"Zhanhao Li","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Zhanhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160392561","display_name":"Shao Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4391012619","display_name":"Shanghai Artificial Intelligence Laboratory","ror":"https://ror.org/03wkvpx79","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4391012619"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Shao","raw_affiliation_strings":["Shanghai Artificial Intelligence Laboratory (shlab.org.cn"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Artificial Intelligence Laboratory (shlab.org.cn","institution_ids":["https://openalex.org/I4391012619"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5160408764","display_name":"Shuyue Hu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4391012619","display_name":"Shanghai Artificial Intelligence Laboratory","ror":"https://ror.org/03wkvpx79","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4391012619"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hu, Shuyue","raw_affiliation_strings":["Shanghai Artificial Intelligence Laboratory (shlab.org.cn"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shanghai Artificial Intelligence Laboratory (shlab.org.cn","institution_ids":["https://openalex.org/I4391012619"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"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":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10203","display_name":"Recommender Systems and Techniques","score":0.576200008392334,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10203","display_name":"Recommender Systems and Techniques","score":0.576200008392334,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.1445000022649765,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.094200000166893,"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/chatbots","display_name":"chatbots","score":0.4880000054836273},{"id":"https://openalex.org/keywords/preference-alignment","display_name":"preference alignment","score":0.5099999904632568}],"concepts":[{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.824400007724762},{"id":"https://openalex.org/C183003079","wikidata":"https://www.wikidata.org/wiki/Q1000371","display_name":"Personalization","level":2,"score":0.7595000267028809},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7174999713897705},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.6492999792098999},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.5425000190734863},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.5145999789237976},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.4900999963283539},{"id":"https://openalex.org/C2779305910","wikidata":"https://www.wikidata.org/wiki/Q5172809","display_name":"Corrective feedback","level":2,"score":0.4758000075817108},{"id":"https://openalex.org/C2777868144","wikidata":"https://www.wikidata.org/wiki/Q7239817","display_name":"Preference elicitation","level":3,"score":0.44510000944137573},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.3910999894142151},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3831000030040741},{"id":"https://openalex.org/C67712803","wikidata":"https://www.wikidata.org/wiki/Q7901853","display_name":"User modeling","level":3,"score":0.3695000112056732},{"id":"https://openalex.org/C2780689630","wikidata":"https://www.wikidata.org/wiki/Q2081815","display_name":"Driving simulator","level":2,"score":0.27730000019073486},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2711000144481659},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2669000029563904}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2611.02667","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2611.02667","pdf_url":"https://arxiv.org/pdf/2611.02667","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2611.02667","doi":"https://doi.org/10.48550/arxiv.2611.02667","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2611.02667","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2611.02667","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2611.02667","pdf_url":"https://arxiv.org/pdf/2611.02667","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7225148442.pdf","grobid_xml":"https://content.openalex.org/works/W7225148442.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7212169225","https://openalex.org/W4416054679","https://openalex.org/W4415330372","https://openalex.org/W7221779681","https://openalex.org/W4407590599","https://openalex.org/W7162415723","https://openalex.org/W4391709257","https://openalex.org/W6948294992","https://openalex.org/W6948435175","https://openalex.org/W7221771754"],"abstract_inverted_index":{"Personalizing":[0],"large":[1],"language":[2],"model":[3,70],"assistants":[4],"requires":[5],"adapting":[6],"their":[7],"responses":[8,96,154],"to":[9,37,66,155,199,204],"individual":[10,130],"users,":[11,172],"including":[12],"the":[13,89,101,111,125,151,160,180,191,218],"selection":[14],"of":[15,20,24,92,163],"content,":[16],"reasoning":[17],"approach,":[18],"level":[19],"detail,":[21],"and":[22,41,94,109,189,201,208,222],"form":[23],"presentation.":[25],"Such":[26],"preferences":[27],"are":[28],"often":[29],"latent,":[30],"becoming":[31],"visible":[32],"only":[33],"when":[34,115],"users":[35],"respond":[36],"an":[38,129],"initial":[39,93],"answer":[40],"request":[42],"revisions.":[43],"We":[44,73,158],"investigate":[45],"whether":[46],"these":[47,214],"feedback":[48],"trajectories":[49],"can":[50,232],"be":[51,156],"converted":[52],"into":[53],"a":[54,84,106,117,146],"reusable":[55,235],"personalization":[56],"signal":[57],"without":[58],"explicit":[59],"preference":[60,113,177,220],"profiles":[61],"or":[62,69,137],"labels,":[63],"prompt-time":[64],"access":[65],"interaction":[67,230],"histories,":[68],"parameter":[71],"updates.":[72],"introduce":[74],"UPSHIFT":[75,164],"(User":[76],"Preference":[77],"Steering":[78],"from":[79,197,202,217],"Human-AI":[80],"Interaction":[81],"Feedback":[82],"Traces),":[83],"training-free":[85],"method":[86,121],"that":[87,141,148,213,229],"compares":[88],"hidden":[90],"representations":[91],"feedback-conditioned":[95],"within":[97],"each":[98],"session,":[99],"averages":[100],"resulting":[102],"trajectory":[103],"displacements":[104],"across":[105,143],"user's":[107],"sessions,":[108],"injects":[110],"aggregated":[112],"direction":[114,221],"answering":[116],"new":[118],"query.":[119],"The":[120],"is":[122],"motivated":[123],"by":[124],"implicit-preference":[126],"hypothesis:":[127],"although":[128],"conversational":[131],"transition":[132],"may":[133],"reflect":[134],"topic":[135],"progression":[136],"task-specific":[138],"correction,":[139],"changes":[140],"recur":[142],"sessions":[144],"retain":[145],"component":[147],"captures":[149],"how":[150],"user":[152],"wants":[153],"adjusted.":[157],"validate":[159],"preference-representation":[161],"capability":[162],"on":[165,183],"PrefEval.":[166],"In":[167],"evaluations":[168],"with":[169],"40":[170],"simulated":[171],"every":[173],"tested":[174],"configuration":[175,193],"improves":[176],"alignment":[178],"over":[179],"unsteered":[181],"baseline":[182],"held-out":[184],"topics":[185],"for":[186,237],"both":[187],"Qwen3-8B":[188],"Qwen3-32B;":[190],"best":[192],"increases":[194],"macro-averaged":[195],"Elo":[196],"1364.0":[198],"1605.2":[200],"1341.9":[203],"1591.8,":[205],"respectively.":[206],"Matched-user":[207],"positive-versus-negative":[209],"controls":[210],"further":[211],"demonstrate":[212],"gains":[215],"arise":[216],"user-specific":[219],"its":[223],"learned":[224],"orientation.":[225],"These":[226],"results":[227],"show":[228],"histories":[231],"provide":[233],"compact,":[234],"supervision":[236],"lightweight":[238],"personalization.":[239]},"counts_by_year":[],"updated_date":"2026-11-06T12:05:46.655746","created_date":"2026-11-04T00:00:00"}