{"id":"https://openalex.org/W4387964173","doi":"https://doi.org/10.48550/arxiv.2310.16834","title":"Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution","display_name":"Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution","publication_year":2023,"publication_date":"2023-10-25","ids":{"openalex":"https://openalex.org/W4387964173","doi":"https://doi.org/10.48550/arxiv.2310.16834"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2310.16834","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2310.16834","pdf_url":"https://arxiv.org/pdf/2310.16834","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2310.16834","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5005937976","display_name":"Aaron Lou","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lou, Aaron","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103112611","display_name":"Chenlin Meng","orcid":"https://orcid.org/0009-0001-9661-0131","observed_orcids":["https://orcid.org/0009-0001-9661-0131","https://orcid.org/0000-0001-8185-7716"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Meng, Chenlin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5091179481","display_name":"Stefano Ermon","orcid":"https://orcid.org/0000-0003-0039-2887","observed_orcids":["https://orcid.org/0000-0003-0039-2887"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ermon, Stefano","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":true,"cited_by_count":5,"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/T10028","display_name":"Topic Modeling","score":0.7006999850273132,"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/T10028","display_name":"Topic Modeling","score":0.7006999850273132,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.1964000016450882,"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"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.03519999980926514,"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/discrete-diffusion-models","display_name":"discrete diffusion models","score":0.9940000176429749},{"id":"https://openalex.org/keywords/score-matching","display_name":"score matching","score":0.9679999947547913},{"id":"https://openalex.org/keywords/discrete-data-generation","display_name":"discrete data generation","score":0.6190000176429749},{"id":"https://openalex.org/keywords/language-modeling","display_name":"language modeling","score":0.6710000038146973},{"id":"https://openalex.org/keywords/perplexity-reduction","display_name":"perplexity reduction","score":0.7080000042915344},{"id":"https://openalex.org/keywords/controllable-text-generation","display_name":"controllable text generation","score":0.6850000023841858},{"id":"https://openalex.org/keywords/infilling","display_name":"infilling","score":0.7089999914169312},{"id":"https://openalex.org/keywords/distribution-ratio","display_name":"distribution ratio","score":0.335999995470047},{"id":"https://openalex.org/keywords/text-generation","display_name":"text generation","score":0.3449999988079071}],"concepts":[{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.7496225833892822},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6153874397277832},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5523149967193604},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.5460978150367737},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.43874692916870117},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.4360136091709137},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.34476739168167114},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.33574506640434265},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33461132645606995},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.27409857511520386},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.2694532573223114},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.10154294967651367}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2310.16834","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2310.16834","pdf_url":"https://arxiv.org/pdf/2310.16834","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2310.16834","doi":"https://doi.org/10.48550/arxiv.2310.16834","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2310.16834","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":"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2310.16834","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2310.16834","pdf_url":"https://arxiv.org/pdf/2310.16834","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G1432373144","display_name":null,"funder_award_id":"W911NF-21-1-","funder_id":"https://openalex.org/F4320338281","funder_display_name":"Army Research Office"},{"id":"https://openalex.org/G2781775652","display_name":"CAREER: Modeling and Inference for Large Scale Spatio-Temporal Data","funder_award_id":"1651565","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G5704118319","display_name":null,"funder_award_id":"W911NF-21-1-0125","funder_id":"https://openalex.org/F4320338281","funder_display_name":"Army Research Office"},{"id":"https://openalex.org/G7452299184","display_name":null,"funder_award_id":"W911NF","funder_id":"https://openalex.org/F4320338281","funder_display_name":"Army Research Office"},{"id":"https://openalex.org/G8876996369","display_name":null,"funder_award_id":"N00014","funder_id":"https://openalex.org/F4320337345","funder_display_name":"Office of Naval Research"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320337345","display_name":"Office of Naval Research","ror":"https://ror.org/00rk2pe57"},{"id":"https://openalex.org/F4320338281","display_name":"Army Research Office","ror":"https://ror.org/05epdh915"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4387964173.pdf","grobid_xml":"https://content.openalex.org/works/W4387964173.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4399151647","https://openalex.org/W4399795196","https://openalex.org/W4415346723","https://openalex.org/W4416105349","https://openalex.org/W4320843382","https://openalex.org/W4385571393","https://openalex.org/W4407170293","https://openalex.org/W4415065813","https://openalex.org/W4392862012","https://openalex.org/W7154221019"],"abstract_inverted_index":{"Despite":[0],"their":[1],"groundbreaking":[2],"performance":[3],"for":[4],"many":[5],"generative":[6,141],"modeling":[7,95],"tasks,":[8],"diffusion":[9,24,76,105],"models":[10,25,90],"have":[11,42],"fallen":[12],"short":[13],"on":[14,27,92],"discrete":[15,40,69,75],"data":[16],"domains":[17],"such":[18],"as":[19],"natural":[20],"language.":[21],"Crucially,":[22],"standard":[23,93],"rely":[26],"the":[28,45],"well-established":[29],"theory":[30],"of":[31],"score":[32,58,66],"matching,":[33],"but":[34],"efforts":[35],"to":[36,39,68,73,123,172],"generalize":[37],"this":[38,50,54],"structures":[41],"not":[43],"yielded":[44],"same":[46],"empirical":[47],"gains.":[48],"In":[49],"work,":[51],"we":[52,83],"bridge":[53],"gap":[55],"by":[56,109],"proposing":[57],"entropy,":[59],"a":[60],"novel":[61],"loss":[62],"that":[63],"naturally":[64],"extends":[65],"matching":[67],"spaces,":[70],"integrates":[71],"seamlessly":[72],"build":[74],"models,":[77,116],"and":[78,111,149,158],"significantly":[79],"boosts":[80],"performance.":[81],"Experimentally,":[82],"test":[84],"our":[85],"Score":[86],"Entropy":[87],"Discrete":[88],"Diffusion":[89],"(SEDD)":[91],"language":[94,104],"tasks.":[96],"For":[97],"comparable":[98],"model":[99],"sizes,":[100],"SEDD":[101,126],"beats":[102],"existing":[103],"paradigms":[106],"(reducing":[107],"perplexity":[108,142],"$25$-$75$\\%)":[110],"is":[112],"competitive":[113],"with":[114,153],"autoregressive":[115,124],"in":[117],"particular":[118],"outperforming":[119],"GPT-2.":[120],"Furthermore,":[121],"compared":[122],"mdoels,":[125],"generates":[127],"faithful":[128],"text":[129],"without":[130],"requiring":[131],"distribution":[132],"annealing":[133],"techniques":[134],"like":[135],"temperature":[136],"scaling":[137],"(around":[138],"$6$-$8\\times$":[139],"better":[140],"than":[143],"un-annealed":[144],"GPT-2),":[145],"can":[146],"trade":[147],"compute":[148],"quality":[150,152,165],"(similar":[151],"$32\\times$":[154],"fewer":[155],"network":[156],"evaluations),":[157],"enables":[159],"controllable":[160],"infilling":[161],"(matching":[162],"nucleus":[163],"sampling":[164],"while":[166],"enabling":[167],"other":[168],"strategies":[169],"besides":[170],"left":[171],"right":[173],"prompting).":[174]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":3}],"updated_date":"2026-10-29T09:56:32.521200","created_date":"2023-10-27T00:00:00"}