{"id":"https://openalex.org/W7131076213","doi":"https://doi.org/10.48550/arxiv.2602.18176","title":"Improving Sampling for Masked Diffusion Models via Information Gain","display_name":"Improving Sampling for Masked Diffusion Models via Information Gain","publication_year":2026,"publication_date":"2026-02-20","ids":{"openalex":"https://openalex.org/W7131076213","doi":"https://doi.org/10.48550/arxiv.2602.18176"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.18176","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":"Article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5121620202","display_name":"Kaisen Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Yang, Kaisen","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5116167297","display_name":"Jayden Teoh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Teoh, Jayden","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126645442","display_name":"Kaicheng Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Kaicheng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126606110","display_name":"Yitong Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Yitong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5126626827","display_name":"Alex Lamb","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lamb, Alex","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5121620202"],"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.7074000239372253,"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"}},"topics":[{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.7074000239372253,"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":0.08860000222921371,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.016899999231100082,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.8158000111579895},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.6747000217437744},{"id":"https://openalex.org/keywords/flexibility","display_name":"Flexibility (engineering)","score":0.5981000065803528},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.5932000279426575},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.5590999722480774},{"id":"https://openalex.org/keywords/trace","display_name":"TRACE (psycholinguistics)","score":0.4875999987125397},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.4586000144481659},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.4374000132083893},{"id":"https://openalex.org/keywords/autoregressive-model","display_name":"Autoregressive model","score":0.43059998750686646}],"concepts":[{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.8158000111579895},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7174000144004822},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.6747000217437744},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.5981000065803528},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.5932000279426575},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.5590999722480774},{"id":"https://openalex.org/C75291252","wikidata":"https://www.wikidata.org/wiki/Q1315756","display_name":"TRACE (psycholinguistics)","level":2,"score":0.4875999987125397},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4586000144481659},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.4374000132083893},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.43059998750686646},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.4296000003814697},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.38999998569488525},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36160001158714294},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.3474999964237213},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3366999924182892},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.33340001106262207},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.3203999996185303},{"id":"https://openalex.org/C189645446","wikidata":"https://www.wikidata.org/wiki/Q350865","display_name":"Mirroring","level":2,"score":0.31859999895095825},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.3001999855041504},{"id":"https://openalex.org/C51823790","wikidata":"https://www.wikidata.org/wiki/Q504353","display_name":"Greedy algorithm","level":2,"score":0.28540000319480896},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.27459999918937683},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.274399995803833},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.2676999866962433},{"id":"https://openalex.org/C193969084","wikidata":"https://www.wikidata.org/wiki/Q7452500","display_name":"Sequential decoding","level":4,"score":0.26429998874664307},{"id":"https://openalex.org/C100776233","wikidata":"https://www.wikidata.org/wiki/Q2532492","display_name":"Bridge (graph theory)","level":2,"score":0.25279998779296875},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2513999938964844}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.18176","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":"Article"},{"id":"doi:10.48550/arxiv.2602.18176","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.18176","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":"article"}],"best_oa_location":{"id":"pmh:doi:10.48550/arxiv.2602.18176","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":"Article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.6229740381240845,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Masked":[0],"Diffusion":[1],"Models":[2],"(MDMs)":[3],"offer":[4],"greater":[5],"flexibility":[6],"in":[7,152,162],"decoding":[8,58,87,108],"order":[9],"than":[10],"autoregressive":[11],"models":[12],"but":[13],"require":[14],"careful":[15],"planning":[16],"to":[17,34,65,175],"achieve":[18],"high-quality":[19],"generation.":[20],"Existing":[21],"samplers":[22,142],"typically":[23],"adopt":[24],"greedy":[25],"heuristics,":[26],"prioritizing":[27],"positions":[28],"with":[29,114],"the":[30,53,77,103,178],"highest":[31],"local":[32],"certainty":[33],"decode":[35],"at":[36,190],"each":[37],"step.":[38],"Through":[39],"failure":[40],"case":[41],"analysis,":[42],"we":[43,101],"identify":[44],"a":[45,86,106,149,159,182],"fundamental":[46],"limitation":[47],"of":[48,56,80],"this":[49,99],"approach:":[50],"it":[51,147,169],"neglects":[52],"downstream":[54],"impact":[55],"current":[57],"choices":[59],"on":[60,155,166],"subsequent":[61],"steps":[62],"and":[63,126,132,158],"fails":[64],"minimize":[66],"cumulative":[67,171],"uncertainty.":[68],"In":[69],"particular,":[70],"these":[71],"methods":[72],"do":[73],"not":[74],"fully":[75],"exploit":[76],"non-causal":[78],"nature":[79],"MDMs,":[81],"which":[82],"enables":[83],"evaluating":[84],"how":[85],"decision":[88],"reshapes":[89],"token":[90],"probabilities/uncertainty":[91],"across":[92,123],"all":[93],"remaining":[94],"masked":[95,119],"positions.":[96],"To":[97],"bridge":[98],"gap,":[100],"propose":[102],"Info-Gain":[104,137],"Sampler,":[105],"principled":[107],"framework":[109],"that":[110,136],"balances":[111],"immediate":[112],"uncertainty":[113,172],"information":[115],"gain":[116],"over":[117],"future":[118],"tokens.":[120],"Extensive":[121],"evaluations":[122],"diverse":[124],"architectures":[125],"tasks":[127,157,168],"(reasoning,":[128],"coding,":[129],"creative":[130,163],"writing,":[131],"image":[133],"generation)":[134],"demonstrate":[135],"Sampler":[138],"consistently":[139],"outperforms":[140],"existing":[141],"for":[143],"MDMs.":[144],"For":[145],"instance,":[146],"achieves":[148],"3.6%":[150],"improvement":[151],"average":[153],"accuracy":[154],"reasoning":[156,167],"63.1%":[160],"win-rate":[161],"writing.":[164],"Notably,":[165],"reduces":[170],"from":[173],"78.4":[174],"48.6,":[176],"outperforming":[177],"best":[179],"baseline":[180],"by":[181],"large":[183],"margin.":[184],"The":[185],"code":[186],"will":[187],"be":[188],"available":[189],"https://github.com/yks23/Information-Gain-Sampler.":[191]},"counts_by_year":[],"updated_date":"2026-04-04T16:13:02.066488","created_date":"2026-02-24T00:00:00"}