{"id":"https://openalex.org/W7102952122","doi":"https://doi.org/10.48550/arxiv.2510.25766","title":"Decomposition-Enhanced Training for Post-Hoc Attributions In Language Models","display_name":"Decomposition-Enhanced Training for Post-Hoc Attributions In Language Models","publication_year":2025,"publication_date":"2025-10-29","ids":{"openalex":"https://openalex.org/W7102952122","doi":"https://doi.org/10.48550/arxiv.2510.25766"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2510.25766","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.25766","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2510.25766","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Balasubramanian, Sriram","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Balasubramanian, Sriram","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Basu, Samyadeep","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Basu, Samyadeep","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Goswami, Koustava","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Goswami, Koustava","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Rossi, Ryan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rossi, Ryan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Manjunatha, Varun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Manjunatha, Varun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Santhosh, Roshan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Santhosh, Roshan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Zhang, Ruiyi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Ruiyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Feizi, Soheil","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feizi, Soheil","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Lipka, Nedim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lipka, Nedim","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":true,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.867900013923645,"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/T10028","display_name":"Topic Modeling","score":0.867900013923645,"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/T13910","display_name":"Computational and Text Analysis Methods","score":0.025800000876188278,"subfield":{"id":"https://openalex.org/subfields/3300","display_name":"General Social Sciences"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.011300000362098217,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/attribution","display_name":"Attribution","score":0.8769999742507935},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.6118999719619751},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.5601999759674072},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.43799999356269836},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.41179999709129333},{"id":"https://openalex.org/keywords/authorship-attribution","display_name":"Authorship attribution","score":0.39989998936653137},{"id":"https://openalex.org/keywords/computational-linguistics","display_name":"Computational linguistics","score":0.38589999079704285},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.3790999948978424}],"concepts":[{"id":"https://openalex.org/C143299363","wikidata":"https://www.wikidata.org/wiki/Q900584","display_name":"Attribution","level":2,"score":0.8769999742507935},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.6118999719619751},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5845999717712402},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.5601999759674072},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5424000024795532},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5223000049591064},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.43799999356269836},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.41179999709129333},{"id":"https://openalex.org/C3020202489","wikidata":"https://www.wikidata.org/wiki/Q2032038","display_name":"Authorship attribution","level":2,"score":0.39989998936653137},{"id":"https://openalex.org/C155092808","wikidata":"https://www.wikidata.org/wiki/Q182557","display_name":"Computational linguistics","level":2,"score":0.38589999079704285},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.3790999948978424},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.3677999973297119},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.36500000953674316},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.3352000117301941},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.33410000801086426},{"id":"https://openalex.org/C2779439875","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Natural language understanding","level":3,"score":0.32910001277923584},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.32170000672340393},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.31859999895095825},{"id":"https://openalex.org/C64357122","wikidata":"https://www.wikidata.org/wiki/Q1149766","display_name":"Causality (physics)","level":2,"score":0.31779998540878296},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.3102000057697296},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.28360000252723694},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.2827000021934509},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.27709999680519104},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C197947376","wikidata":"https://www.wikidata.org/wiki/Q5155608","display_name":"Comparability","level":2,"score":0.25619998574256897},{"id":"https://openalex.org/C2778571376","wikidata":"https://www.wikidata.org/wiki/Q1355821","display_name":"Frontier","level":2,"score":0.25519999861717224}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2510.25766","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.25766","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2510.25766","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.25766","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,"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"},"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":{"Large":[0],"language":[1],"models":[2,76,96],"(LLMs)":[3],"are":[4,61],"increasingly":[5],"used":[6],"for":[7,18,26],"long-document":[8],"question":[9],"answering,":[10],"where":[11,37,59],"reliable":[12],"attribution":[13,22,51,146],"to":[14,68,77,97],"sources":[15],"is":[16],"critical":[17],"trust.":[19],"Existing":[20],"post-hoc":[21,50],"methods":[23,150],"work":[24],"well":[25],"extractive":[27],"QA":[28,112],"but":[29],"struggle":[30],"in":[31],"multi-hop,":[32],"abstractive,":[33],"and":[34,121,125,141,151],"semi-extractive":[35],"settings,":[36],"answers":[38,60],"synthesize":[39],"information":[40],"across":[41],"passages.":[42],"To":[43],"address":[44],"these":[45],"challenges,":[46],"we":[47,88],"argue":[48],"that":[49,74,94],"can":[52],"be":[53],"reframed":[54],"as":[55,101],"a":[56,91,107,118,128],"reasoning":[57,103],"problem,":[58],"decomposed":[62],"into":[63],"constituent":[64],"units,":[65],"each":[66],"tied":[67],"specific":[69],"context.":[70],"We":[71,105],"first":[72],"show":[73],"prompting":[75],"generate":[78],"such":[79],"decompositions":[80,100,116],"alongside":[81],"attributions":[82],"improves":[83,145],"performance.":[84],"Building":[85],"on":[86],"this,":[87],"introduce":[89],"DecompTune,":[90],"post-training":[92],"method":[93],"teaches":[95],"produce":[98],"answer":[99],"intermediate":[102],"steps.":[104],"curate":[106],"diverse":[108],"dataset":[109],"of":[110],"complex":[111],"tasks,":[113],"annotated":[114],"with":[115,134],"by":[117],"strong":[119],"LLM,":[120],"post-train":[122],"Qwen-2.5":[123],"(7B":[124],"14B)":[126],"using":[127],"two-stage":[129],"SFT":[130],"+":[131],"GRPO":[132],"pipeline":[133],"task-specific":[135],"curated":[136],"rewards.":[137],"Across":[138],"extensive":[139],"experiments":[140],"ablations,":[142],"DecompTune":[143],"substantially":[144],"quality,":[147],"outperforming":[148],"prior":[149],"matching":[152],"or":[153],"exceeding":[154],"state-of-the-art":[155],"frontier":[156],"models.":[157]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2025-10-31T00:00:00"}