{"id":"https://openalex.org/W7162780065","doi":"https://doi.org/10.48550/arxiv.2605.29076","title":"Structured Prompt Optimization Meets Reinforcement Learning for Global and Local Interpretability over Complex Text","display_name":"Structured Prompt Optimization Meets Reinforcement Learning for Global and Local Interpretability over Complex Text","publication_year":2026,"publication_date":"2026-05-27","ids":{"openalex":"https://openalex.org/W7162780065","doi":"https://doi.org/10.48550/arxiv.2605.29076"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.29076","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29076","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":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.2605.29076","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5040442343","display_name":"Tianyang Zhou","orcid":"https://orcid.org/0000-0002-8392-3972"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Tianyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137385303","display_name":"Wenbo Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Wenbo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006034181","display_name":"Pierre Jinghong Liang","orcid":"https://orcid.org/0000-0002-9246-3269"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liang, Pierre Jinghong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137397666","display_name":"Leman Akoglu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Akoglu, Leman","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":0.8428999781608582,"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":0.8428999781608582,"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":0.03449999913573265,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.0284000001847744,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/interpretability","display_name":"Interpretability","score":0.859499990940094},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.7263000011444092},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6004999876022339},{"id":"https://openalex.org/keywords/modular-design","display_name":"Modular design","score":0.5404000282287598},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.48809999227523804},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.453900009393692},{"id":"https://openalex.org/keywords/face","display_name":"Face (sociological concept)","score":0.44339999556541443}],"concepts":[{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.859499990940094},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.744700014591217},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7263000011444092},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6378999948501587},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6004999876022339},{"id":"https://openalex.org/C101468663","wikidata":"https://www.wikidata.org/wiki/Q1620158","display_name":"Modular design","level":2,"score":0.5404000282287598},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.48809999227523804},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.453900009393692},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.44339999556541443},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41690000891685486},{"id":"https://openalex.org/C2779439875","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Natural language understanding","level":3,"score":0.37549999356269836},{"id":"https://openalex.org/C2776639384","wikidata":"https://www.wikidata.org/wiki/Q840396","display_name":"Ideal (ethics)","level":2,"score":0.35269999504089355},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.34200000762939453},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2924000024795532},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.2732999920845032},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.26409998536109924},{"id":"https://openalex.org/C44210515","wikidata":"https://www.wikidata.org/wiki/Q16968978","display_name":"Bespoke","level":2,"score":0.26100000739097595},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.2590999901294708},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.25029999017715454}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.29076","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29076","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":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.2605.29076","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29076","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.5295195579528809}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"LLMs":[0],"have":[1],"advanced":[2],"text":[3,23],"classification,":[4],"yet":[5],"existing":[6,129],"paradigms":[7,130],"face":[8],"a":[9,54,65,76,81,106,116],"trade-off:":[10],"supervised":[11],"(label":[12],"only)":[13],"fine-tuning":[14],"is":[15],"scalable":[16],"but":[17,36],"offers":[18,33],"limited":[19],"reasoning":[20,73,87,113],"on":[21],"complex":[22],"and":[24,40,84,138],"lacks":[25],"broader":[26],"model":[27],"transparency,":[28],"while":[29,125],"discrete":[30],"prompt":[31],"optimization":[32],"human-readable":[34],"instructions":[35],"struggles":[37],"with":[38,48,109,141],"performance":[39,137],"scalability.":[41],"We":[42],"introduce":[43],"eXTC":[44,99],"(eXplainable":[45],"Text":[46],"Classifier)":[47],"three":[49],"progressive":[50],"stages:":[51],"(1)":[52],"learning":[53],"Standard":[55],"Operating":[56],"Procedure":[57],"(SOP,":[58],"or":[59],"rulebook)":[60],"in":[61,134],"natural":[62],"language":[63],"via":[64,93,105],"new":[66],"Structured":[67],"Prompt":[68],"Optimization":[69],"algorithm;":[70],"(2)":[71],"SOP-grounded":[72],"distillation":[74],"from":[75],"large":[77],"teacher":[78],"LLM":[79],"into":[80],"compact":[82,107],"LM;":[83],"(3)":[85],"expanding":[86],"capabilities":[88],"beyond":[89],"the":[90],"initial":[91],"SOP":[92],"reinforcement":[94],"learning.":[95],"This":[96],"design":[97],"enables":[98],"to":[100],"provide":[101],"(i)":[102],"fast":[103],"inference":[104],"LM,":[108],"(ii)":[110],"inference-time":[111],"local":[112],"traces,":[114],"alongside":[115],"global,":[117],"modular":[118],"explanation":[119,139],"of":[120],"its":[121],"learned":[122],"domain":[123],"rules,":[124],"(iii)":[126],"significantly":[127],"outperforming":[128],"across":[131],"diverse":[132],"benchmarks":[133],"both":[135],"classification":[136],"quality,":[140],"stage-by-stage":[142],"gains.":[143]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-30T00:00:00"}