{"id":"https://openalex.org/W7166884110","doi":"https://doi.org/10.18653/v1/2026.acl-short.67","title":"T\u22c6: Progressive Block Scaling for Masked Diffusion Language Models Through Trajectory Aware Reinforcement Learning","display_name":"T\u22c6: Progressive Block Scaling for Masked Diffusion Language Models Through Trajectory Aware Reinforcement Learning","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166884110","doi":"https://doi.org/10.18653/v1/2026.acl-short.67"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-short.67","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-short.67","pdf_url":"https://aclanthology.org/2026.acl-short.67.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-short.67.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5123329644","display_name":"Hanchen Xia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hanchen Xia","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123310416","display_name":"Baoyou Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Baoyou Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123330383","display_name":"Yutang Ge","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yutang Ge","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068045654","display_name":"Guojiang Zhao","orcid":"https://orcid.org/0000-0001-7166-7089"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guojiang Zhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139753225","display_name":"Siyu Zhu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Siyu Zhu","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.87240664,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"808","last_page":"816"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12031","display_name":"Speech and dialogue systems","score":0.2565999925136566,"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/T12031","display_name":"Speech and dialogue systems","score":0.2565999925136566,"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":0.10580000281333923,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.06520000100135803,"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/trajectory","display_name":"Trajectory","score":0.683899998664856},{"id":"https://openalex.org/keywords/block","display_name":"Block (permutation group theory)","score":0.609499990940094},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.5379999876022339},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.4634000062942505},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.36250001192092896},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.3188999891281128},{"id":"https://openalex.org/keywords/diffusion","display_name":"Diffusion","score":0.3156999945640564}],"concepts":[{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.683899998664856},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6625000238418579},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.609499990940094},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5379999876022339},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5131000280380249},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.4634000062942505},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.385699987411499},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.36250001192092896},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.3188999891281128},{"id":"https://openalex.org/C69357855","wikidata":"https://www.wikidata.org/wiki/Q163214","display_name":"Diffusion","level":2,"score":0.3156999945640564},{"id":"https://openalex.org/C2993597952","wikidata":"https://www.wikidata.org/wiki/Q108123644","display_name":"Block model","level":2,"score":0.3093000054359436},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.2865999937057495},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.2621999979019165},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.25859999656677246},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2542000114917755},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.2517000138759613}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-short.67","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-short.67","pdf_url":"https://aclanthology.org/2026.acl-short.67.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-short.67","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-short.67","pdf_url":"https://aclanthology.org/2026.acl-short.67.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10","score":0.4080498218536377}],"awards":[],"funders":[{"id":"https://openalex.org/F4320309612","display_name":"Natural Science Foundation of Shanghai","ror":null},{"id":"https://openalex.org/F4320325413","display_name":"Shanghai Municipal Commission of Economy and Informatization","ror":"https://ror.org/04166ws88"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166884110.pdf","grobid_xml":"https://content.openalex.org/works/W7166884110.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"We":[0],"present":[1],"T":[2,24,55],"\u22c6":[3,25,56],",":[4],"a":[5,81],"simple":[6],"TRACERL-based":[7],"curriculum":[8],"for":[9],"progressive":[10],"block-size":[11],"scaling":[12],"in":[13],"masked":[14],"diffusion":[15],"language":[16],"models":[17],"(MDMs).Starting":[18],"from":[19],"an":[20],"AR-initialized":[21],"small-block":[22],"MDM,":[23],"gradually":[26],"increases":[27],"the":[28,33,73],"block":[29],"size":[30],"while":[31,91],"re-optimizing":[32],"denoising":[34],"policy":[35,75],"at":[36],"each":[37],"stage,":[38],"enabling":[39],"higher-parallelism":[40],"decoding":[41],"with":[42],"limited":[43],"degradation":[44],"on":[45],"math":[46],"reasoning":[47],"benchmarks.Across":[48],"two":[49],"SDAR":[50],"scales":[51],"and":[52,62],"three":[53],"benchmarks,":[54],"consistently":[57],"outperforms":[58],"direct":[59],"large-block":[60],"TRACERL":[61],"is":[63],"substantially":[64],"more":[65],"stable":[66],"during":[67],"training.Our":[68],"schedule":[69],"analysis":[70],"suggests":[71],"that":[72],"learned":[74],"does":[76],"not":[77],"simply":[78],"revert":[79],"to":[80],"strictly":[82],"left-to-right":[83],"order;":[84],"instead,":[85],"it":[86],"retains":[87],"block-size-specific":[88],"nonmonotone":[89],"updates":[90],"improving":[92],"accuracy.":[93]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}