Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion Article Swipe
YOU?
·
· 2024
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2407.01392
This paper presents Diffusion Forcing, a new training paradigm where a diffusion model is trained to denoise a set of tokens with independent per-token noise levels. We apply Diffusion Forcing to sequence generative modeling by training a causal next-token prediction model to generate one or several future tokens without fully diffusing past ones. Our approach is shown to combine the strengths of next-token prediction models, such as variable-length generation, with the strengths of full-sequence diffusion models, such as the ability to guide sampling to desirable trajectories. Our method offers a range of additional capabilities, such as (1) rolling-out sequences of continuous tokens, such as video, with lengths past the training horizon, where baselines diverge and (2) new sampling and guiding schemes that uniquely profit from Diffusion Forcing's variable-horizon and causal architecture, and which lead to marked performance gains in decision-making and planning tasks. In addition to its empirical success, our method is proven to optimize a variational lower bound on the likelihoods of all subsequences of tokens drawn from the true joint distribution. Project website: https://boyuan.space/diffusion-forcing
Related Topics
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2407.01392
- https://arxiv.org/pdf/2407.01392
- OA Status
- green
- Cited By
- 3
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4400373517
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4400373517Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2407.01392Digital Object Identifier
- Title
-
Diffusion Forcing: Next-token Prediction Meets Full-Sequence DiffusionWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2024Year of publication
- Publication date
-
2024-07-01Full publication date if available
- Authors
-
Boyuan Chen, Diego Martí Monsó, Yilun Du, Max Simchowitz, Russ Tedrake, Vincent SitzmannList of authors in order
- Landing page
-
https://arxiv.org/abs/2407.01392Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2407.01392Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2407.01392Direct OA link when available
- Concepts
-
Diffusion, Forcing (mathematics), Security token, Sequence (biology), Computer science, Algorithm, Mathematics, Physics, Mathematical analysis, Computer network, Chemistry, Thermodynamics, BiochemistryTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
3Total citation count in OpenAlex
- Citations by year (recent)
-
2025: 3Per-year citation counts (last 5 years)
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4400373517 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2407.01392 |
| ids.doi | https://doi.org/10.48550/arxiv.2407.01392 |
| ids.openalex | https://openalex.org/W4400373517 |
| fwci | |
| type | preprint |
| title | Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T10751 |
| topics[0].field.id | https://openalex.org/fields/13 |
| topics[0].field.display_name | Biochemistry, Genetics and Molecular Biology |
| topics[0].score | 0.8431000113487244 |
| topics[0].domain.id | https://openalex.org/domains/1 |
| topics[0].domain.display_name | Life Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/1311 |
| topics[0].subfield.display_name | Genetics |
| topics[0].display_name | Forensic and Genetic Research |
| topics[1].id | https://openalex.org/T11242 |
| topics[1].field.id | https://openalex.org/fields/25 |
| topics[1].field.display_name | Materials Science |
| topics[1].score | 0.8177000284194946 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/2505 |
| topics[1].subfield.display_name | Materials Chemistry |
| topics[1].display_name | Nuclear Materials and Properties |
| topics[2].id | https://openalex.org/T10346 |
| topics[2].field.id | https://openalex.org/fields/31 |
| topics[2].field.display_name | Physics and Astronomy |
| topics[2].score | 0.816100001335144 |
| topics[2].domain.id | https://openalex.org/domains/3 |
| topics[2].domain.display_name | Physical Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/3106 |
| topics[2].subfield.display_name | Nuclear and High Energy Physics |
| topics[2].display_name | Magnetic confinement fusion research |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C69357855 |
| concepts[0].level | 2 |
| concepts[0].score | 0.8000619411468506 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q163214 |
| concepts[0].display_name | Diffusion |
| concepts[1].id | https://openalex.org/C197115733 |
| concepts[1].level | 2 |
| concepts[1].score | 0.706277072429657 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q1003136 |
| concepts[1].display_name | Forcing (mathematics) |
| concepts[2].id | https://openalex.org/C48145219 |
| concepts[2].level | 2 |
| concepts[2].score | 0.6818171739578247 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q1335365 |
| concepts[2].display_name | Security token |
| concepts[3].id | https://openalex.org/C2778112365 |
| concepts[3].level | 2 |
| concepts[3].score | 0.6723659038543701 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q3511065 |
| concepts[3].display_name | Sequence (biology) |
| concepts[4].id | https://openalex.org/C41008148 |
| concepts[4].level | 0 |
| concepts[4].score | 0.43616771697998047 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[4].display_name | Computer science |
| concepts[5].id | https://openalex.org/C11413529 |
| concepts[5].level | 1 |
| concepts[5].score | 0.35247403383255005 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q8366 |
| concepts[5].display_name | Algorithm |
| concepts[6].id | https://openalex.org/C33923547 |
| concepts[6].level | 0 |
| concepts[6].score | 0.2967511713504791 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q395 |
| concepts[6].display_name | Mathematics |
| concepts[7].id | https://openalex.org/C121332964 |
| concepts[7].level | 0 |
| concepts[7].score | 0.18672117590904236 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q413 |
| concepts[7].display_name | Physics |
| concepts[8].id | https://openalex.org/C134306372 |
| concepts[8].level | 1 |
| concepts[8].score | 0.16279712319374084 |
| concepts[8].wikidata | https://www.wikidata.org/wiki/Q7754 |
| concepts[8].display_name | Mathematical analysis |
| concepts[9].id | https://openalex.org/C31258907 |
| concepts[9].level | 1 |
| concepts[9].score | 0.11264905333518982 |
| concepts[9].wikidata | https://www.wikidata.org/wiki/Q1301371 |
| concepts[9].display_name | Computer network |
| concepts[10].id | https://openalex.org/C185592680 |
| concepts[10].level | 0 |
| concepts[10].score | 0.10980203747749329 |
| concepts[10].wikidata | https://www.wikidata.org/wiki/Q2329 |
| concepts[10].display_name | Chemistry |
| concepts[11].id | https://openalex.org/C97355855 |
| concepts[11].level | 1 |
| concepts[11].score | 0.08837887644767761 |
| concepts[11].wikidata | https://www.wikidata.org/wiki/Q11473 |
| concepts[11].display_name | Thermodynamics |
| concepts[12].id | https://openalex.org/C55493867 |
| concepts[12].level | 1 |
| concepts[12].score | 0.0 |
| concepts[12].wikidata | https://www.wikidata.org/wiki/Q7094 |
| concepts[12].display_name | Biochemistry |
| keywords[0].id | https://openalex.org/keywords/diffusion |
| keywords[0].score | 0.8000619411468506 |
| keywords[0].display_name | Diffusion |
| keywords[1].id | https://openalex.org/keywords/forcing |
| keywords[1].score | 0.706277072429657 |
| keywords[1].display_name | Forcing (mathematics) |
| keywords[2].id | https://openalex.org/keywords/security-token |
| keywords[2].score | 0.6818171739578247 |
| keywords[2].display_name | Security token |
| keywords[3].id | https://openalex.org/keywords/sequence |
| keywords[3].score | 0.6723659038543701 |
| keywords[3].display_name | Sequence (biology) |
| keywords[4].id | https://openalex.org/keywords/computer-science |
| keywords[4].score | 0.43616771697998047 |
| keywords[4].display_name | Computer science |
| keywords[5].id | https://openalex.org/keywords/algorithm |
| keywords[5].score | 0.35247403383255005 |
| keywords[5].display_name | Algorithm |
| keywords[6].id | https://openalex.org/keywords/mathematics |
| keywords[6].score | 0.2967511713504791 |
| keywords[6].display_name | Mathematics |
| keywords[7].id | https://openalex.org/keywords/physics |
| keywords[7].score | 0.18672117590904236 |
| keywords[7].display_name | Physics |
| keywords[8].id | https://openalex.org/keywords/mathematical-analysis |
| keywords[8].score | 0.16279712319374084 |
| keywords[8].display_name | Mathematical analysis |
| keywords[9].id | https://openalex.org/keywords/computer-network |
| keywords[9].score | 0.11264905333518982 |
| keywords[9].display_name | Computer network |
| keywords[10].id | https://openalex.org/keywords/chemistry |
| keywords[10].score | 0.10980203747749329 |
| keywords[10].display_name | Chemistry |
| keywords[11].id | https://openalex.org/keywords/thermodynamics |
| keywords[11].score | 0.08837887644767761 |
| keywords[11].display_name | Thermodynamics |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2407.01392 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400194 |
| locations[0].source.issn | |
| locations[0].source.type | repository |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | arXiv (Cornell University) |
| locations[0].source.host_organization | https://openalex.org/I205783295 |
| locations[0].source.host_organization_name | Cornell University |
| locations[0].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[0].license | |
| locations[0].pdf_url | https://arxiv.org/pdf/2407.01392 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2407.01392 |
| locations[1].id | doi:10.48550/arxiv.2407.01392 |
| locations[1].is_oa | True |
| locations[1].source.id | https://openalex.org/S4306400194 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | True |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | arXiv (Cornell University) |
| locations[1].source.host_organization | https://openalex.org/I205783295 |
| locations[1].source.host_organization_name | Cornell University |
| locations[1].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[1].license | cc-by |
| locations[1].pdf_url | |
| locations[1].version | |
| locations[1].raw_type | article |
| locations[1].license_id | https://openalex.org/licenses/cc-by |
| locations[1].is_accepted | False |
| locations[1].is_published | |
| locations[1].raw_source_name | |
| locations[1].landing_page_url | https://doi.org/10.48550/arxiv.2407.01392 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5103094528 |
| authorships[0].author.orcid | https://orcid.org/0009-0003-4012-1659 |
| authorships[0].author.display_name | Boyuan Chen |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Chen, Boyuan |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5100029705 |
| authorships[1].author.orcid | |
| authorships[1].author.display_name | Diego Martí Monsó |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Monso, Diego Marti |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5101182304 |
| authorships[2].author.orcid | |
| authorships[2].author.display_name | Yilun Du |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Du, Yilun |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5037154191 |
| authorships[3].author.orcid | https://orcid.org/0000-0001-9900-1238 |
| authorships[3].author.display_name | Max Simchowitz |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Simchowitz, Max |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5074291890 |
| authorships[4].author.orcid | |
| authorships[4].author.display_name | Russ Tedrake |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | Tedrake, Russ |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5016061808 |
| authorships[5].author.orcid | https://orcid.org/0000-0002-0107-5704 |
| authorships[5].author.display_name | Vincent Sitzmann |
| authorships[5].author_position | last |
| authorships[5].raw_author_name | Sitzmann, Vincent |
| authorships[5].is_corresponding | False |
| has_content.pdf | False |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2407.01392 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-10T00:00:00 |
| display_name | Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T10751 |
| primary_topic.field.id | https://openalex.org/fields/13 |
| primary_topic.field.display_name | Biochemistry, Genetics and Molecular Biology |
| primary_topic.score | 0.8431000113487244 |
| primary_topic.domain.id | https://openalex.org/domains/1 |
| primary_topic.domain.display_name | Life Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/1311 |
| primary_topic.subfield.display_name | Genetics |
| primary_topic.display_name | Forensic and Genetic Research |
| related_works | https://openalex.org/W4388335561, https://openalex.org/W2970530566, https://openalex.org/W4288261899, https://openalex.org/W4307309205, https://openalex.org/W2967478618, https://openalex.org/W2051487156, https://openalex.org/W4385009901, https://openalex.org/W4385572700, https://openalex.org/W2997152889, https://openalex.org/W2073681303 |
| cited_by_count | 3 |
| counts_by_year[0].year | 2025 |
| counts_by_year[0].cited_by_count | 3 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2407.01392 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400194 |
| best_oa_location.source.issn | |
| best_oa_location.source.type | repository |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | arXiv (Cornell University) |
| best_oa_location.source.host_organization | https://openalex.org/I205783295 |
| best_oa_location.source.host_organization_name | Cornell University |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| best_oa_location.license | |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2407.01392 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | http://arxiv.org/abs/2407.01392 |
| primary_location.id | pmh:oai:arXiv.org:2407.01392 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400194 |
| primary_location.source.issn | |
| primary_location.source.type | repository |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | arXiv (Cornell University) |
| primary_location.source.host_organization | https://openalex.org/I205783295 |
| primary_location.source.host_organization_name | Cornell University |
| primary_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| primary_location.license | |
| primary_location.pdf_url | https://arxiv.org/pdf/2407.01392 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2407.01392 |
| publication_date | 2024-07-01 |
| publication_year | 2024 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 5, 10, 17, 36, 89, 155 |
| abstract_inverted_index.In | 143 |
| abstract_inverted_index.We | 26 |
| abstract_inverted_index.as | 66, 77, 95, 103 |
| abstract_inverted_index.by | 34 |
| abstract_inverted_index.in | 138 |
| abstract_inverted_index.is | 13, 55, 151 |
| abstract_inverted_index.of | 19, 61, 72, 91, 99, 162, 165 |
| abstract_inverted_index.on | 159 |
| abstract_inverted_index.or | 44 |
| abstract_inverted_index.to | 15, 30, 41, 57, 80, 83, 134, 145, 153 |
| abstract_inverted_index.(1) | 96 |
| abstract_inverted_index.(2) | 115 |
| abstract_inverted_index.Our | 53, 86 |
| abstract_inverted_index.all | 163 |
| abstract_inverted_index.and | 114, 118, 128, 131, 140 |
| abstract_inverted_index.its | 146 |
| abstract_inverted_index.new | 6, 116 |
| abstract_inverted_index.one | 43 |
| abstract_inverted_index.our | 149 |
| abstract_inverted_index.set | 18 |
| abstract_inverted_index.the | 59, 70, 78, 108, 160, 169 |
| abstract_inverted_index.This | 0 |
| abstract_inverted_index.from | 124, 168 |
| abstract_inverted_index.lead | 133 |
| abstract_inverted_index.past | 51, 107 |
| abstract_inverted_index.such | 65, 76, 94, 102 |
| abstract_inverted_index.that | 121 |
| abstract_inverted_index.true | 170 |
| abstract_inverted_index.with | 21, 69, 105 |
| abstract_inverted_index.apply | 27 |
| abstract_inverted_index.bound | 158 |
| abstract_inverted_index.drawn | 167 |
| abstract_inverted_index.fully | 49 |
| abstract_inverted_index.gains | 137 |
| abstract_inverted_index.guide | 81 |
| abstract_inverted_index.joint | 171 |
| abstract_inverted_index.lower | 157 |
| abstract_inverted_index.model | 12, 40 |
| abstract_inverted_index.noise | 24 |
| abstract_inverted_index.ones. | 52 |
| abstract_inverted_index.paper | 1 |
| abstract_inverted_index.range | 90 |
| abstract_inverted_index.shown | 56 |
| abstract_inverted_index.where | 9, 111 |
| abstract_inverted_index.which | 132 |
| abstract_inverted_index.causal | 37, 129 |
| abstract_inverted_index.future | 46 |
| abstract_inverted_index.marked | 135 |
| abstract_inverted_index.method | 87, 150 |
| abstract_inverted_index.offers | 88 |
| abstract_inverted_index.profit | 123 |
| abstract_inverted_index.proven | 152 |
| abstract_inverted_index.tasks. | 142 |
| abstract_inverted_index.tokens | 20, 47, 166 |
| abstract_inverted_index.video, | 104 |
| abstract_inverted_index.Forcing | 29 |
| abstract_inverted_index.Project | 173 |
| abstract_inverted_index.ability | 79 |
| abstract_inverted_index.combine | 58 |
| abstract_inverted_index.denoise | 16 |
| abstract_inverted_index.diverge | 113 |
| abstract_inverted_index.guiding | 119 |
| abstract_inverted_index.lengths | 106 |
| abstract_inverted_index.levels. | 25 |
| abstract_inverted_index.models, | 64, 75 |
| abstract_inverted_index.schemes | 120 |
| abstract_inverted_index.several | 45 |
| abstract_inverted_index.tokens, | 101 |
| abstract_inverted_index.trained | 14 |
| abstract_inverted_index.without | 48 |
| abstract_inverted_index.Forcing, | 4 |
| abstract_inverted_index.addition | 144 |
| abstract_inverted_index.approach | 54 |
| abstract_inverted_index.generate | 42 |
| abstract_inverted_index.horizon, | 110 |
| abstract_inverted_index.modeling | 33 |
| abstract_inverted_index.optimize | 154 |
| abstract_inverted_index.paradigm | 8 |
| abstract_inverted_index.planning | 141 |
| abstract_inverted_index.presents | 2 |
| abstract_inverted_index.sampling | 82, 117 |
| abstract_inverted_index.sequence | 31 |
| abstract_inverted_index.success, | 148 |
| abstract_inverted_index.training | 7, 35, 109 |
| abstract_inverted_index.uniquely | 122 |
| abstract_inverted_index.website: | 174 |
| abstract_inverted_index.Diffusion | 3, 28, 125 |
| abstract_inverted_index.Forcing's | 126 |
| abstract_inverted_index.baselines | 112 |
| abstract_inverted_index.desirable | 84 |
| abstract_inverted_index.diffusing | 50 |
| abstract_inverted_index.diffusion | 11, 74 |
| abstract_inverted_index.empirical | 147 |
| abstract_inverted_index.per-token | 23 |
| abstract_inverted_index.sequences | 98 |
| abstract_inverted_index.strengths | 60, 71 |
| abstract_inverted_index.additional | 92 |
| abstract_inverted_index.continuous | 100 |
| abstract_inverted_index.generative | 32 |
| abstract_inverted_index.next-token | 38, 62 |
| abstract_inverted_index.prediction | 39, 63 |
| abstract_inverted_index.generation, | 68 |
| abstract_inverted_index.independent | 22 |
| abstract_inverted_index.likelihoods | 161 |
| abstract_inverted_index.performance | 136 |
| abstract_inverted_index.rolling-out | 97 |
| abstract_inverted_index.variational | 156 |
| abstract_inverted_index.subsequences | 164 |
| abstract_inverted_index.architecture, | 130 |
| abstract_inverted_index.capabilities, | 93 |
| abstract_inverted_index.distribution. | 172 |
| abstract_inverted_index.full-sequence | 73 |
| abstract_inverted_index.trajectories. | 85 |
| abstract_inverted_index.decision-making | 139 |
| abstract_inverted_index.variable-length | 67 |
| abstract_inverted_index.variable-horizon | 127 |
| abstract_inverted_index.https://boyuan.space/diffusion-forcing | 175 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 6 |
| citation_normalized_percentile |