{"id":"https://openalex.org/W7160841808","doi":"https://doi.org/10.48550/arxiv.2605.07693","title":"Toward Better Geometric Representations for Molecule Generative Models","display_name":"Toward Better Geometric Representations for Molecule Generative Models","publication_year":2026,"publication_date":"2026-05-08","ids":{"openalex":"https://openalex.org/W7160841808","doi":"https://doi.org/10.48550/arxiv.2605.07693"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.07693","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07693","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":"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.2605.07693","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135884929","display_name":"Shaoheng Yan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yan, Shaoheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126252602","display_name":"Zian Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Zian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135870304","display_name":"Cai Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Cai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055586473","display_name":"Qiaojing Huang","orcid":"https://orcid.org/0000-0002-8303-5879"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Qiaojing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135868116","display_name":"Kai Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Kai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135871189","display_name":"Muhan Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Muhan","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":0.6434000134468079,"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":0.6434000134468079,"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":0.09880000352859497,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.03849999979138374,"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/representation","display_name":"Representation (politics)","score":0.7110000252723694},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.5856000185012817},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.5767999887466431},{"id":"https://openalex.org/keywords/generator","display_name":"Generator (circuit theory)","score":0.45339998602867126},{"id":"https://openalex.org/keywords/decoupling","display_name":"Decoupling (probability)","score":0.44850000739097595},{"id":"https://openalex.org/keywords/equivalence","display_name":"Equivalence (formal languages)","score":0.3797000050544739},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.365200012922287}],"concepts":[{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.7110000252723694},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6349999904632568},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.5856000185012817},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.5767999887466431},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5353000164031982},{"id":"https://openalex.org/C2780992000","wikidata":"https://www.wikidata.org/wiki/Q17016113","display_name":"Generator (circuit theory)","level":3,"score":0.45339998602867126},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.44850000739097595},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.3797000050544739},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.365200012922287},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.35530000925064087},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.33500000834465027},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.3303000032901764},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3294000029563904},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.31790000200271606},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3174999952316284},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2809000015258789},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2678000032901764},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2540999948978424}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.07693","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07693","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":"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.2605.07693","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07693","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":"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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Geometric":[0],"representation-conditioned":[1,99],"molecule":[2,10,18,26,33,96,128,175,182,192],"generation":[3,19,43,100,176,183],"provides":[4],"an":[5],"effective":[6],"paradigm":[7,227],"that":[8,70,90,118,131,148],"decouples":[9],"representation":[11,60,111,138,144],"modeling":[12],"from":[13,122],"structure":[14],"generation.":[15,165],"By":[16],"decoupling":[17],"into":[20],"two":[21],"stages-first":[22],"generating":[23,30],"a":[24,31,88,110,127,136,142,224],"meaningful":[25],"representation,":[27],"and":[28,39,73,140,164,190,205],"then":[29],"3D":[32],"conditioned":[34],"on":[35,179],"this":[36,83],"representation-the":[37],"efficiency":[38],"quality":[40,57],"of":[41,58,95,170],"the":[42,56,59,78,93,123,133,151,159,168,180,211],"process":[44],"can":[45],"be":[46],"significantly":[47],"enhanced.":[48],"However,":[49],"its":[50],"effectiveness":[51,169],"is":[52],"fundamentally":[53],"limited":[54],"by":[55],"space:":[61],"pretrained":[62,124],"molecular":[63,229],"encoders,":[64],"such":[65],"as":[66,223],"UniMol,":[67],"produce":[68],"representations":[69,97,121],"are":[71],"non-smooth":[72],"not":[74],"fully":[75],"exploited":[76],"during":[77,115],"generative":[79,116,218],"training":[80,219],"process.":[81],"In":[82,102],"work,":[84],"we":[85],"propose":[86],"LENSEs,":[87],"framework":[89],"better":[91],"exploits":[92],"potential":[94,225],"in":[98,135],"methods.":[101,197],"particular,":[103],"LENSEs":[104,186],"introduces":[105],"three":[106],"complementary":[107],"mechanisms:":[108],"(1)":[109],"head,":[112],"simultaneously":[113],"trained":[114],"tasks,":[117],"extracts":[119],"multi-level":[120],"encoder;":[125],"(2)":[126],"perceptual":[129],"loss":[130,147],"optimizes":[132],"generator":[134],"semantic-informative":[137],"space;":[139],"(3)":[141],"node-level":[143],"alignment":[145,221],"(REPA)":[146],"explicitly":[149],"aligns":[150],"generator's":[152],"hidden":[153],"states":[154],"with":[155,220],"encoder":[156],"representations,":[157,216],"reducing":[158],"semantic":[160],"gap":[161],"between":[162],"pretraining":[163,226],"We":[166],"demonstrate":[167,210],"these":[171],"improvements":[172],"through":[173,200],"extensive":[174],"tasks.":[177],"Specifically,":[178],"challenging":[181],"dataset":[184],"GEOM-DRUG,":[185],"achieves":[187],"97.28%":[188],"validity":[189],"98.51%":[191],"stability,":[193],"surpassing":[194],"existing":[195],"advanced":[196],"Further":[198],"analyses":[199],"Lipschitz":[201],"constant":[202],"reduction":[203],"(4.6x)":[204],"QM9":[206],"probing":[207],"tasks":[208],"also":[209],"smoother,":[212],"more":[213],"informative":[214],"refined":[215],"establishing":[217],"objectives":[222],"for":[228],"encoders.":[230]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-12T00:00:00"}