{"id":"https://openalex.org/W7204174679","doi":"https://doi.org/10.48550/arxiv.2608.21897","title":"From Solver Feedback to Faithful Plans: Multi-Role Reinforcement Learning for Symbolic Planning","display_name":"From Solver Feedback to Faithful Plans: Multi-Role Reinforcement Learning for Symbolic Planning","publication_year":2026,"publication_date":"2026-08-22","ids":{"openalex":"https://openalex.org/W7204174679","doi":"https://doi.org/10.48550/arxiv.2608.21897"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.21897","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21897","pdf_url":"https://arxiv.org/pdf/2608.21897","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.21897","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5121349031","display_name":"Chenghao Zhang","orcid":"https://orcid.org/0009-0008-0277-2166","observed_orcids":["https://orcid.org/0009-0008-0277-2166"]},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhang, Chenghao","raw_affiliation_strings":["University of Pennsylvania"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Pennsylvania","institution_ids":["https://openalex.org/I79576946"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027836225","display_name":"Yikai Mao","orcid":"https://orcid.org/0009-0008-4182-3173","observed_orcids":["https://orcid.org/0009-0008-4182-3173"]},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mao, Yikai","raw_affiliation_strings":["University of Pennsylvania"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Pennsylvania","institution_ids":["https://openalex.org/I79576946"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023586151","display_name":"Haoyu Gao","orcid":"https://orcid.org/0000-0002-1407-5180","observed_orcids":["https://orcid.org/0000-0002-1407-5180"]},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gao, Haoyu","raw_affiliation_strings":["Georgia Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135939873","display_name":"Saisai Hu","orcid":"https://orcid.org/0009-0007-4099-4037","observed_orcids":["https://orcid.org/0009-0007-4099-4037"]},"institutions":[{"id":"https://openalex.org/I126863827","display_name":"Pace University","ror":"https://ror.org/047p7y759","country_code":"US","type":"education","lineage":["https://openalex.org/I126863827"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hu, Saisai","raw_affiliation_strings":["Pace University "],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pace University ","institution_ids":["https://openalex.org/I126863827"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013618931","display_name":"Yuxi Cheng","orcid":"https://orcid.org/0009-0000-9393-4993","observed_orcids":["https://orcid.org/0009-0000-9393-4993"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cheng, Yuxi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153446183","display_name":"Shanqi Liu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Shanqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5023802054","display_name":"Dan Roth","orcid":"https://orcid.org/0009-0002-1447-5173","observed_orcids":["https://orcid.org/0009-0002-1447-5173"]},"institutions":[{"id":"https://openalex.org/I79576946","display_name":"University of Pennsylvania","ror":"https://ror.org/00b30xv10","country_code":"US","type":"education","lineage":["https://openalex.org/I79576946"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Roth, Dan","raw_affiliation_strings":["University of Pennsylvania"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Pennsylvania","institution_ids":["https://openalex.org/I79576946"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.9779999852180481,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.9779999852180481,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.009800000116229057,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.005799999926239252,"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/symbolic-planning","display_name":"symbolic planning","score":0.8479999899864197},{"id":"https://openalex.org/keywords/pddl","display_name":"PDDL","score":0.6890000104904175},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8529999852180481},{"id":"https://openalex.org/keywords/semantic-drift","display_name":"semantic drift","score":0.527999997138977},{"id":"https://openalex.org/keywords/plan-verification","display_name":"plan verification","score":0.5149999856948853},{"id":"https://openalex.org/keywords/plan-repair","display_name":"plan repair","score":0.6980000138282776}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8210999965667725},{"id":"https://openalex.org/C160145156","wikidata":"https://www.wikidata.org/wiki/Q778586","display_name":"Executable","level":2,"score":0.7910000085830688},{"id":"https://openalex.org/C2778770139","wikidata":"https://www.wikidata.org/wiki/Q1966904","display_name":"Solver","level":2,"score":0.7627999782562256},{"id":"https://openalex.org/C164155591","wikidata":"https://www.wikidata.org/wiki/Q2067766","display_name":"Satisfiability modulo theories","level":2,"score":0.6182000041007996},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6029999852180481},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.5734000205993652},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.573199987411499},{"id":"https://openalex.org/C114073186","wikidata":"https://www.wikidata.org/wiki/Q2631895","display_name":"Automated planning and scheduling","level":2,"score":0.5647000074386597},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5169000029563904},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.39800000190734863},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3529999852180481},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3513999879360199},{"id":"https://openalex.org/C110251889","wikidata":"https://www.wikidata.org/wiki/Q1569697","display_name":"Model checking","level":2,"score":0.335999995470047},{"id":"https://openalex.org/C61423126","wikidata":"https://www.wikidata.org/wiki/Q187432","display_name":"Scripting language","level":2,"score":0.3271999955177307},{"id":"https://openalex.org/C3019612716","wikidata":"https://www.wikidata.org/wiki/Q730920","display_name":"Problem solver","level":2,"score":0.31150001287460327},{"id":"https://openalex.org/C34388435","wikidata":"https://www.wikidata.org/wiki/Q2267362","display_name":"Bounded function","level":2,"score":0.30570000410079956},{"id":"https://openalex.org/C207609745","wikidata":"https://www.wikidata.org/wiki/Q4944086","display_name":"Bootstrapping (finance)","level":2,"score":0.30300000309944153},{"id":"https://openalex.org/C179603123","wikidata":"https://www.wikidata.org/wiki/Q1941921","display_name":"Modeling language","level":3,"score":0.28619998693466187},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.2741999924182892},{"id":"https://openalex.org/C2779639559","wikidata":"https://www.wikidata.org/wiki/Q7661178","display_name":"Symbolic execution","level":3,"score":0.2648000121116638},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.2635999917984009},{"id":"https://openalex.org/C2776095079","wikidata":"https://www.wikidata.org/wiki/Q489538","display_name":"The Symbolic","level":2,"score":0.2533000111579895}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.21897","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21897","pdf_url":"https://arxiv.org/pdf/2608.21897","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.2608.21897","doi":"https://doi.org/10.48550/arxiv.2608.21897","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.21897","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:2608.21897","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21897","pdf_url":"https://arxiv.org/pdf/2608.21897","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":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204174679.pdf","grobid_xml":"https://content.openalex.org/works/W7204174679.grobid-xml"},"referenced_works_count":16,"referenced_works":["https://openalex.org/W3030163527","https://openalex.org/W4320559489","https://openalex.org/W4322718191","https://openalex.org/W4327810158","https://openalex.org/W4389520747","https://openalex.org/W4391631327","https://openalex.org/W4393160795","https://openalex.org/W4399511704","https://openalex.org/W4403758464","https://openalex.org/W4412886640","https://openalex.org/W4414281281","https://openalex.org/W4416010183","https://openalex.org/W4416037109","https://openalex.org/W7120088840","https://openalex.org/W7133236450","https://openalex.org/W7140615325"],"related_works":["https://openalex.org/W4405032655","https://openalex.org/W7215996179","https://openalex.org/W7203689856","https://openalex.org/W7166658681","https://openalex.org/W7212128920","https://openalex.org/W4407309453","https://openalex.org/W4414455564","https://openalex.org/W7133188479","https://openalex.org/W4416259053","https://openalex.org/W7111601702"],"abstract_inverted_index":{"Reliable":[0],"planning":[1,131],"requires":[2],"converting":[3],"natural-language":[4],"instructions":[5],"into":[6,118],"executable":[7],"symbolic":[8,130],"specifications,":[9,73],"yet":[10],"large":[11],"language":[12,55],"models":[13],"remain":[14],"brittle":[15],"without":[16,41],"costly":[17],"PDDL":[18,72],"annotations":[19],"and":[20,62,67,81,105,121,127],"may":[21],"exploit":[22],"solver":[23,39,116],"success":[24,94],"in":[25],"semantically":[26],"unfaithful":[27],"ways.":[28],"We":[29,44],"study":[30],"how":[31],"to":[32,99,109],"learn":[33],"faithful":[34,103,128],"natural-language-to-PDDL":[35],"formalization":[36],"using":[37],"only":[38],"feedback,":[40],"human-written":[42],"demonstrations.":[43],"propose":[45],"a":[46,53,77],"solvergrounded":[47],"multi-role":[48],"reinforcement":[49],"learning":[50],"framework":[51],"where":[52],"single":[54],"model":[56],"acts":[57],"as":[58],"an":[59],"Actor,":[60],"Judge,":[61],"Editor":[63,83],"for":[64,97],"generation,":[65,119],"verification,":[66,120],"repair.":[68],"The":[69],"Actor":[70],"proposes":[71],"the":[74,82],"Judge":[75],"provides":[76],"solver-calibrated":[78],"quality":[79],"signal,":[80],"performs":[84],"bounded":[85],"diagnostic-conditioned":[86],"refinement.":[87],"On":[88],"PlanBench,":[89],"our":[90],"method":[91],"improves":[92],"average":[93],"from":[95],"35.5%":[96],"LLM+P":[98],"70.8%,":[100],"achieves":[101],"66.3%":[102],"success,":[104],"reduces":[106],"semantic":[107],"drift":[108],"6.4%.":[110],"These":[111],"results":[112],"show":[113],"that":[114],"organizing":[115],"feedback":[117],"repair":[122],"roles":[123],"enables":[124],"more":[125],"scalable":[126],"annotation-free":[129]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-26T00:00:00"}