{"id":"https://openalex.org/W7157969354","doi":"https://doi.org/10.1145/3805622.3810780","title":"Efficient Rationale-based Retrieval: On-policy Distillation from Generative Rerankers based on JEPA","display_name":"Efficient Rationale-based Retrieval: On-policy Distillation from Generative Rerankers based on JEPA","publication_year":2026,"publication_date":"2026-06-15","ids":{"openalex":"https://openalex.org/W7157969354","doi":"https://doi.org/10.1145/3805622.3810780"},"language":null,"primary_location":{"id":"doi:10.1145/3805622.3810780","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805622.3810780","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 International Conference on Multimedia Retrieval","raw_type":"proceedings-article"},"type":"article","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3805622.3810780","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134845029","display_name":"Teng Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I4210153393","display_name":"Geely (China)","ror":"https://ror.org/0446d5v35","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153393"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Teng Chen","raw_affiliation_strings":["Geely AI Lab, Ningbo, Zhejiang, China"],"raw_orcid":"https://orcid.org/0009-0000-3604-5313","affiliations":[{"raw_affiliation_string":"Geely AI Lab, Ningbo, Zhejiang, China","institution_ids":["https://openalex.org/I4210153393"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134874967","display_name":"Sheng Xu","orcid":null},"institutions":[{"id":"https://openalex.org/I4210153393","display_name":"Geely (China)","ror":"https://ror.org/0446d5v35","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153393"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Sheng Xu","raw_affiliation_strings":["Geely AI Lab, Ningbo, Zhejiang, China"],"raw_orcid":"https://orcid.org/0009-0006-3352-8882","affiliations":[{"raw_affiliation_string":"Geely AI Lab, Ningbo, Zhejiang, China","institution_ids":["https://openalex.org/I4210153393"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134764127","display_name":"Feixiang Guo","orcid":null},"institutions":[{"id":"https://openalex.org/I4210153393","display_name":"Geely (China)","ror":"https://ror.org/0446d5v35","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153393"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Feixiang Guo","raw_affiliation_strings":["Geely AI Lab, Ningbo, Zhejiang, China"],"raw_orcid":"https://orcid.org/0009-0008-3132-440X","affiliations":[{"raw_affiliation_string":"Geely AI Lab, Ningbo, Zhejiang, China","institution_ids":["https://openalex.org/I4210153393"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134867125","display_name":"Xiaoyu Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210153393","display_name":"Geely (China)","ror":"https://ror.org/0446d5v35","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153393"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaoyu Wang","raw_affiliation_strings":["Geely AI Lab, Ningbo, Zhejiang, China"],"raw_orcid":"https://orcid.org/0009-0003-8617-0616","affiliations":[{"raw_affiliation_string":"Geely AI Lab, Ningbo, Zhejiang, China","institution_ids":["https://openalex.org/I4210153393"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134859907","display_name":"Qingqing Gu","orcid":null},"institutions":[{"id":"https://openalex.org/I4210153393","display_name":"Geely (China)","ror":"https://ror.org/0446d5v35","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153393"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qingqing Gu","raw_affiliation_strings":["Geely AI Lab, Ningbo, Zhejiang, China"],"raw_orcid":"https://orcid.org/0009-0007-6872-8910","affiliations":[{"raw_affiliation_string":"Geely AI Lab, Ningbo, Zhejiang, China","institution_ids":["https://openalex.org/I4210153393"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134871024","display_name":"Hongyan Li","orcid":null},"institutions":[{"id":"https://openalex.org/I4210153393","display_name":"Geely (China)","ror":"https://ror.org/0446d5v35","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153393"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongyan Li","raw_affiliation_strings":["Geely AI Lab, Ningbo, Zhejiang, China"],"raw_orcid":"https://orcid.org/0009-0008-8280-5190","affiliations":[{"raw_affiliation_string":"Geely AI Lab, Ningbo, Zhejiang, China","institution_ids":["https://openalex.org/I4210153393"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5134865127","display_name":"Luo Ji","orcid":null},"institutions":[{"id":"https://openalex.org/I4210153393","display_name":"Geely (China)","ror":"https://ror.org/0446d5v35","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153393"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Luo Ji","raw_affiliation_strings":["Geely AI Lab, Ningbo, Zhejiang, China"],"raw_orcid":"https://orcid.org/0000-0002-2484-5345","affiliations":[{"raw_affiliation_string":"Geely AI Lab, Ningbo, Zhejiang, China","institution_ids":["https://openalex.org/I4210153393"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.61161027,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1721","last_page":"1729"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.29980000853538513,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.29980000853538513,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.26980000734329224,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.1703999936580658,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.7544000148773193},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.4636000096797943},{"id":"https://openalex.org/keywords/relevance","display_name":"Relevance (law)","score":0.4318999946117401},{"id":"https://openalex.org/keywords/contrast","display_name":"Contrast (vision)","score":0.4284999966621399},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.39890000224113464},{"id":"https://openalex.org/keywords/term","display_name":"Term (time)","score":0.3707999885082245},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.3400999903678894},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.3262999951839447}],"concepts":[{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.7544000148773193},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7188000082969666},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4860999882221222},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.4636000096797943},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.4318999946117401},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4300999939441681},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.4284999966621399},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.39890000224113464},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.3707999885082245},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.3400999903678894},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.3262999951839447},{"id":"https://openalex.org/C2778334786","wikidata":"https://www.wikidata.org/wiki/Q1586270","display_name":"Variation (astronomy)","level":2,"score":0.31839999556541443},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.31459999084472656},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.305400013923645},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.2924000024795532},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.29100000858306885},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.27399998903274536},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.26809999346733093},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.2676999866962433},{"id":"https://openalex.org/C24755975","wikidata":"https://www.wikidata.org/wiki/Q4943354","display_name":"Boolean conjunctive query","level":5,"score":0.251800000667572},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.2517000138759613}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3805622.3810780","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805622.3810780","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 International Conference on Multimedia Retrieval","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2604.23336","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2604.23336","pdf_url":"https://arxiv.org/pdf/2604.23336","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"doi:10.1145/3805622.3810780","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805622.3810780","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 International Conference on Multimedia Retrieval","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.7337352633476257,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Unlike":[0],"traditional":[1,208],"fact-based":[2],"retrieval,":[3],"rationale-based":[4,188],"retrieval":[5,209],"typically":[6],"necessitates":[7],"cross-encoding":[8],"of":[9,77,152],"query-document":[10,37],"pairs":[11],"using":[12],"large":[13],"language":[14],"models,":[15],"incurring":[16],"substantial":[17],"computational":[18],"costs.":[19],"To":[20],"address":[21],"this":[22,137],"limitation,":[23],"we":[24],"propose":[25],"Rabtriever,":[26],"which":[27,50,102],"independently":[28],"encodes":[29],"queries":[30],"and":[31,58,111,140,176,193,215],"documents,":[32],"while":[33],"providing":[34],"comparable":[35,218],"cross":[36],"comprehension":[38],"capabilities":[39],"to":[40,55,62,87,155,171,220],"rerankers.":[41],"We":[42,70],"start":[43],"from":[44,200],"training":[45],"a":[46,104,118],"LLM-based":[47],"generative":[48],"reranker,":[49,222],"puts":[51],"the":[52,56,60,64,75,85,89,114,123,127,133,141,149,157,163,168,201,221,226],"document":[53,124,169],"prior":[54],"query":[57,92,115],"prompts":[59],"LLM":[61,109,153],"generate":[63],"relevance":[65],"score":[66],"by":[67],"log":[68],"probabilities.":[69],"then":[71,131],"employ":[72],"it":[73],"as":[74,84,126,212,223,225],"teacher":[76,142],"an":[78,145],"on-policy":[79],"distillation":[80],"framework,":[81],"with":[82,122,144,196,217],"Rabtriever":[83,161,181,203],"student":[86],"reconstruct":[88],"teacher\u2019s":[90,164],"contextual-aware":[91],"embedding.":[93],"The":[94],"Joint-Embedding":[95],"Predictive":[96],"Architecture":[97],"(JEPA)":[98],"paradigm":[99],"is":[100],"adopted,":[101],"integrates":[103],"lightweight,":[105],"trainable":[106],"predictor":[107],"between":[108,136],"layers":[110],"heads,":[112],"projecting":[113],"embedding":[116,125,139],"into":[117],"new":[119],"hidden":[120],"space,":[121],"latent":[128],"vector.":[129],"JEPA":[130],"minimizes":[132],"distribution":[134],"difference":[135],"projected":[138],"embedding,":[143],"auxiliary":[146],"loss":[147],"on":[148,167,207],"reverse":[150],"KL":[151],"logits,":[154],"reshape":[156],"student\u2019s":[158],"logit":[159],"distribution.":[160],"optimizes":[162],"quadratic":[165],"complexity":[166],"length":[170],"linear,":[172],"verified":[173],"both":[174],"theoretically":[175],"empirically.":[177],"Experiments":[178],"show":[179],"that":[180],"outperforms":[182],"different":[183],"retriever":[184,228],"baselines":[185],"across":[186],"diverse":[187],"tasks,":[189],"including":[190],"empathetic":[191],"conversations":[192],"robotic":[194],"manipulations,":[195],"minor":[197],"accuracy":[198],"degradation":[199],"reranker.":[202],"also":[204],"generalizes":[205],"well":[206,224],"benchmarks":[210],"such":[211],"MS":[213],"MARCO":[214],"BEIR,":[216],"performance":[219],"best":[227],"baseline.":[229]},"counts_by_year":[],"updated_date":"2026-06-16T07:37:23.134862","created_date":"2026-04-30T00:00:00"}