{"id":"https://openalex.org/W7137315308","doi":"https://doi.org/10.48550/arxiv.2603.12666","title":"RetroReasoner: A Reasoning LLM for Strategic Retrosynthesis Prediction","display_name":"RetroReasoner: A Reasoning LLM for Strategic Retrosynthesis Prediction","publication_year":2026,"publication_date":"2026-03-13","ids":{"openalex":"https://openalex.org/W7137315308","doi":"https://doi.org/10.48550/arxiv.2603.12666"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.12666","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.12666","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.12666","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5000376670","display_name":"Hanbum Ko","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Ko, Hanbum","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043544720","display_name":"Chanhui Lee","orcid":"https://orcid.org/0009-0003-2092-3670"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Chanhui","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129585020","display_name":"Ye Rin Kim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Ye Rin","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024797101","display_name":"Rodrigo Horm\u00e1zabal","orcid":"https://orcid.org/0000-0003-3744-5401"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hormazabal, Rodrigo","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129499070","display_name":"Sehui Han","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Sehui","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129482463","display_name":"Sungbin Lim","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lim, Sungbin","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5087294413","display_name":"Sungwoong Kim","orcid":"https://orcid.org/0000-0002-2676-9454"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kim, Sungwoong","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5000376670"],"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/T11948","display_name":"Machine Learning in Materials Science","score":0.8033000230789185,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.8033000230789185,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10211","display_name":"Computational Drug Discovery Methods","score":0.024900000542402267,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T11281","display_name":"Asymmetric Hydrogenation and Catalysis","score":0.02459999918937683,"subfield":{"id":"https://openalex.org/subfields/1604","display_name":"Inorganic Chemistry"},"field":{"id":"https://openalex.org/fields/16","display_name":"Chemistry"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/retrosynthetic-analysis","display_name":"Retrosynthetic analysis","score":0.710099995136261},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.6654000282287598},{"id":"https://openalex.org/keywords/disconnection","display_name":"Disconnection","score":0.5759999752044678},{"id":"https://openalex.org/keywords/product","display_name":"Product (mathematics)","score":0.5572999715805054},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.45489999651908875},{"id":"https://openalex.org/keywords/core","display_name":"Core (optical fiber)","score":0.37400001287460327}],"concepts":[{"id":"https://openalex.org/C42437451","wikidata":"https://www.wikidata.org/wiki/Q902227","display_name":"Retrosynthetic analysis","level":3,"score":0.710099995136261},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6654000282287598},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.621399998664856},{"id":"https://openalex.org/C2780371621","wikidata":"https://www.wikidata.org/wiki/Q3709910","display_name":"Disconnection","level":2,"score":0.5759999752044678},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5738000273704529},{"id":"https://openalex.org/C90673727","wikidata":"https://www.wikidata.org/wiki/Q901718","display_name":"Product (mathematics)","level":2,"score":0.5572999715805054},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4629000127315521},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.45489999651908875},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.37400001287460327},{"id":"https://openalex.org/C539667460","wikidata":"https://www.wikidata.org/wiki/Q2414942","display_name":"Management science","level":1,"score":0.32829999923706055},{"id":"https://openalex.org/C19351080","wikidata":"https://www.wikidata.org/wiki/Q1395034","display_name":"New product development","level":2,"score":0.3215999901294708},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.29280000925064087},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2802000045776367},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.26269999146461487},{"id":"https://openalex.org/C2779231336","wikidata":"https://www.wikidata.org/wiki/Q7534724","display_name":"Sketch","level":2,"score":0.2621000111103058},{"id":"https://openalex.org/C2778348673","wikidata":"https://www.wikidata.org/wiki/Q739302","display_name":"Production (economics)","level":2,"score":0.2522999942302704},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.25189998745918274}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.12666","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.12666","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.12666","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.12666","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":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Retrosynthesis":[0],"prediction":[1],"is":[2,31,97],"a":[3,15,22,60,87,113,130,141,171],"core":[4],"task":[5],"in":[6,40,179],"organic":[7],"synthesis":[8,143],"that":[9,71,91,115,161],"aims":[10],"to":[11,74],"predict":[12,52],"reactants":[13,53,137],"for":[14],"given":[16],"product":[17,62,152],"molecule.":[18],"Traditionally,":[19],"chemists":[20],"select":[21],"plausible":[23],"bond":[24],"disconnection":[25,118],"and":[26,33,104,145],"derive":[27],"corresponding":[28],"reactants,":[29],"which":[30],"time-consuming":[32],"requires":[34],"substantial":[35],"expertise.":[36],"While":[37],"recent":[38],"advancements":[39],"molecular":[41],"large":[42],"language":[43],"models":[44],"(LLMs)":[45],"have":[46],"made":[47],"progress,":[48],"many":[49],"methods":[50],"either":[51],"without":[54],"strategic":[55,94],"reasoning":[56,89],"or":[57],"conduct":[58],"only":[59,164],"generic":[61],"analysis,":[63],"rather":[64],"than":[65],"reason":[66],"explicitly":[67],"about":[68],"bond-disconnection":[69],"strategies":[70],"logically":[72],"lead":[73],"the":[75,124,150,154],"choice":[76],"of":[77,126,174],"specific":[78],"reactants.":[79],"To":[80],"overcome":[81],"these":[82],"limitations,":[83],"we":[84,110,128],"propose":[85],"RetroReasoner,":[86],"retrosynthetic":[88],"model":[90],"leverages":[92],"chemists'":[93],"thinking.":[95],"RetroReasoner":[96,162],"trained":[98],"using":[99],"both":[100],"supervised":[101],"fine-tuning":[102],"(SFT)":[103],"reinforcement":[105],"learning":[106],"(RL).":[107],"For":[108],"SFT,":[109],"introduce":[111],"SyntheticRetro,":[112],"framework":[114],"generates":[116,170],"structured":[117],"rationales":[119],"alongside":[120],"reactant":[121,176],"predictions.":[122],"In":[123],"case":[125],"RL,":[127],"apply":[129],"round-trip":[131],"accuracy":[132],"as":[133],"reward,":[134],"where":[135],"predicted":[136],"are":[138,147],"passed":[139],"through":[140],"forward":[142],"model,":[144],"predictions":[146],"rewarded":[148],"when":[149],"forward-predicted":[151],"matches":[153],"original":[155],"input":[156],"product.":[157],"Experimental":[158],"results":[159],"show":[160],"not":[163],"outperforms":[165],"prior":[166],"baselines":[167],"but":[168],"also":[169],"broader":[172],"range":[173],"feasible":[175],"proposals,":[177],"particularly":[178],"handling":[180],"more":[181],"challenging":[182],"reaction":[183],"instances.":[184]},"counts_by_year":[],"updated_date":"2026-03-17T07:05:13.627479","created_date":"2026-03-17T00:00:00"}