{"id":"https://openalex.org/W7133543199","doi":"https://doi.org/10.48550/arxiv.2603.02215","title":"RxnNano:Training Compact LLMs for Chemical Reaction and Retrosynthesis Prediction via Hierarchical Curriculum Learning","display_name":"RxnNano:Training Compact LLMs for Chemical Reaction and Retrosynthesis Prediction via Hierarchical Curriculum Learning","publication_year":2026,"publication_date":"2026-02-10","ids":{"openalex":"https://openalex.org/W7133543199","doi":"https://doi.org/10.48550/arxiv.2603.02215"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.02215","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02215","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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.2603.02215","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5128038052","display_name":"Ran Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Ran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128125380","display_name":"Shimin Di","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Di, Shimin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128103523","display_name":"Haowei LI","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"LI, Haowei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128085260","display_name":"Luanshi Bu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bu, Luanshi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128040865","display_name":"Jiachuan Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Jiachuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128103331","display_name":"Wangze Ni","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ni, Wangze","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5128041423","display_name":"Lei Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Lei","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/T11948","display_name":"Machine Learning in Materials Science","score":0.9261999726295471,"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":0.9261999726295471,"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":0.05050000175833702,"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/T12254","display_name":"Machine Learning in Bioinformatics","score":0.0032999999821186066,"subfield":{"id":"https://openalex.org/subfields/1312","display_name":"Molecular Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/intuition","display_name":"Intuition","score":0.4027000069618225},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.40139999985694885},{"id":"https://openalex.org/keywords/retrosynthetic-analysis","display_name":"Retrosynthetic analysis","score":0.39010000228881836},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.37790000438690186},{"id":"https://openalex.org/keywords/knowledge-representation-and-reasoning","display_name":"Knowledge representation and reasoning","score":0.3589000105857849},{"id":"https://openalex.org/keywords/domain-knowledge","display_name":"Domain knowledge","score":0.3569999933242798},{"id":"https://openalex.org/keywords/bridging","display_name":"Bridging (networking)","score":0.3488999903202057}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5548999905586243},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46950000524520874},{"id":"https://openalex.org/C132010649","wikidata":"https://www.wikidata.org/wiki/Q189222","display_name":"Intuition","level":2,"score":0.4027000069618225},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.40139999985694885},{"id":"https://openalex.org/C42437451","wikidata":"https://www.wikidata.org/wiki/Q902227","display_name":"Retrosynthetic analysis","level":3,"score":0.39010000228881836},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.37790000438690186},{"id":"https://openalex.org/C161301231","wikidata":"https://www.wikidata.org/wiki/Q3478658","display_name":"Knowledge representation and reasoning","level":2,"score":0.3589000105857849},{"id":"https://openalex.org/C207685749","wikidata":"https://www.wikidata.org/wiki/Q2088941","display_name":"Domain knowledge","level":2,"score":0.3569999933242798},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.3488999903202057},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32359999418258667},{"id":"https://openalex.org/C60048249","wikidata":"https://www.wikidata.org/wiki/Q37437","display_name":"Syntax","level":2,"score":0.30979999899864197},{"id":"https://openalex.org/C43971567","wikidata":"https://www.wikidata.org/wiki/Q3142865","display_name":"Logical reasoning","level":2,"score":0.30649998784065247},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.29600000381469727},{"id":"https://openalex.org/C124223222","wikidata":"https://www.wikidata.org/wiki/Q2281940","display_name":"Chemical process","level":2,"score":0.2892000079154968},{"id":"https://openalex.org/C169900460","wikidata":"https://www.wikidata.org/wiki/Q2200417","display_name":"Cognition","level":2,"score":0.2825999855995178},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.2791000008583069},{"id":"https://openalex.org/C96146094","wikidata":"https://www.wikidata.org/wiki/Q609057","display_name":"Unification","level":2,"score":0.27320000529289246},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.2587999999523163},{"id":"https://openalex.org/C190470478","wikidata":"https://www.wikidata.org/wiki/Q2370229","display_name":"Invariant (physics)","level":2,"score":0.25450000166893005}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.02215","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02215","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.02215","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02215","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"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":{"Chemical":[0,93],"reaction":[1,40,50],"prediction":[2],"is":[3],"pivotal":[4],"for":[5],"accelerating":[6],"drug":[7],"discovery":[8],"and":[9,26,42,53,108,148,176],"synthesis":[10],"planning.":[11],"Despite":[12],"advances":[13],"in":[14,39,65],"data-driven":[15],"models,":[16],"current":[17],"approaches":[18],"are":[19],"hindered":[20],"by":[21],"an":[22],"overemphasis":[23],"on":[24,101,187],"parameter":[25],"dataset":[27],"scaling.":[28],"Some":[29],"methods":[30],"coupled":[31],"with":[32],"evaluation":[33],"techniques":[34],"that":[35,60,80,96,117],"bypass":[36],"fundamental":[37],"challenges":[38],"representation":[41],"fail":[43],"to":[44,127,143,156],"capture":[45],"deep":[46],"chemical":[47,82,104,132],"intuition":[48],"like":[49],"common":[51],"sense":[52],"{topological":[54],"atom":[55],"mapping":[56],"logic}.":[57],"We":[58],"argue":[59],"the":[61,70,119,141,158,161,178],"core":[62],"challenge":[63],"lies":[64],"instilling":[66],"these":[67],"knowledge":[68],"into":[69],"models.":[71],"To":[72],"this":[73],"end,":[74],"we":[75],"propose":[76],"a":[77,91,102,113,182],"unified":[78],"framework":[79],"prioritizes":[81],"understanding":[83],"over":[84],"scale":[85],"through":[86,121],"three":[87],"key":[88],"innovations:":[89],"(1)":[90],"{Latent":[92],"Consistency}":[94],"objective":[95],"models":[97],"reactions":[98],"as":[99],"movements":[100],"continuous":[103],"manifold,":[105],"ensuring":[106],"reversible":[107],"physically":[109],"plausible":[110],"transformations;":[111],"(2)":[112],"{Hierarchical":[114],"Cognitive":[115],"Curriculum}":[116],"trains":[118],"model":[120,142],"progressive":[122],"stages,":[123],"from":[124],"syntax":[125],"mastery":[126],"semantic":[128],"reasoning,":[129],"building":[130],"robust":[131],"intuition;":[133],"(3)":[134],"{Atom-Map":[135],"Permutation":[136],"Invariance":[137],"(AMPI)},":[138],"which":[139],"force":[140],"learn":[144],"invariant":[145],"relational":[146],"topology":[147],"balance":[149],"multi-task":[150],"learning.":[151],"(4)and":[152],"structured":[153],"plan-based":[154],"reasoning":[155],"improve":[157],"performance":[159],"of":[160],"LLMs.":[162],"Our":[163],"compact":[164],"{0.5B-parameter":[165],"model},":[166],"\\textbf{RxnNano}":[167],"significantly":[168],"outperforms":[169],"fine-tuned":[170],"LLMs":[171],"ten":[172],"times":[173],"larger":[174],"(>7B)":[175],"all":[177],"domain":[179],"baselines,":[180],"achieving":[181],"23.5\\%":[183],"Top-1":[184],"accuracy":[185],"improvement":[186],"rigorous":[188],"benchmarks":[189],"without":[190],"test-time":[191],"augmentation.":[192],"https://github.com/rlisml/RxnNano.":[193]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-05T00:00:00"}