DISCRETE TOKEN PROCESSING USING DIFFUSION MODELS
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output sequence of discrete tokens using a diffusion model. In one aspect, a method includes generating, by using the diffusion model, a final latent representation of the sequence of di...
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Patent |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | Strudel, Robin Grathwohl, Will S Tallec, Corentin Sifre, Laurent Du, Yilin Dieleman, Sander Etienne Lea Savinov, Nikolay Leblond, Rémi Ganin, Iaroslav Mensch, Arthur Altché, Florent |
| description | Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output sequence of discrete tokens using a diffusion model. In one aspect, a method includes generating, by using the diffusion model, a final latent representation of the sequence of discrete tokens that includes a determined value for each of a plurality of latent variables; applying a de-embedding matrix to the final latent representation of the output sequence of discrete tokens to generate a de-embedded final latent representation that includes, for each of the plurality of latent variables, a respective numeric score for each discrete token in a vocabulary of multiple discrete tokens; selecting, for each of the plurality of latent variables, a discrete token from among the multiple discrete tokens in the vocabulary that has a highest numeric score; and generating the output sequence of discrete tokens that includes the selected discrete tokens. |
| format | Patent |
| fullrecord | <record><control><sourceid>epo_EVB</sourceid><recordid>TN_cdi_epo_espacenet_US2024119261A1</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>US2024119261A1</sourcerecordid><originalsourceid>FETCH-epo_espacenet_US2024119261A13</originalsourceid><addsrcrecordid>eNrjZDBw8Qx2DnINcVUI8fd29VMICPJ3dg0O9vRzVwgFky6ebm5Alr-fgq-_i6tPMA8Da1piTnEqL5TmZlB2cw1x9tBNLciPTy0uSExOzUstiQ8NNjIwMjE0tDQyM3Q0NCZOFQDegydl</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>patent</recordtype></control><display><type>patent</type><title>DISCRETE TOKEN PROCESSING USING DIFFUSION MODELS</title><source>esp@cenet</source><creator>Strudel, Robin ; Grathwohl, Will S ; Tallec, Corentin ; Sifre, Laurent ; Du, Yilin ; Dieleman, Sander Etienne Lea ; Savinov, Nikolay ; Leblond, Rémi ; Ganin, Iaroslav ; Mensch, Arthur ; Altché, Florent</creator><creatorcontrib>Strudel, Robin ; Grathwohl, Will S ; Tallec, Corentin ; Sifre, Laurent ; Du, Yilin ; Dieleman, Sander Etienne Lea ; Savinov, Nikolay ; Leblond, Rémi ; Ganin, Iaroslav ; Mensch, Arthur ; Altché, Florent</creatorcontrib><description>Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output sequence of discrete tokens using a diffusion model. In one aspect, a method includes generating, by using the diffusion model, a final latent representation of the sequence of discrete tokens that includes a determined value for each of a plurality of latent variables; applying a de-embedding matrix to the final latent representation of the output sequence of discrete tokens to generate a de-embedded final latent representation that includes, for each of the plurality of latent variables, a respective numeric score for each discrete token in a vocabulary of multiple discrete tokens; selecting, for each of the plurality of latent variables, a discrete token from among the multiple discrete tokens in the vocabulary that has a highest numeric score; and generating the output sequence of discrete tokens that includes the selected discrete tokens.</description><language>eng</language><subject>CALCULATING ; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS ; COMPUTING ; COUNTING ; PHYSICS</subject><creationdate>2024</creationdate><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240411&DB=EPODOC&CC=US&NR=2024119261A1$$EHTML$$P50$$Gepo$$Hfree_for_read</linktohtml><link.rule.ids>230,308,776,881,25543,76293</link.rule.ids><linktorsrc>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240411&DB=EPODOC&CC=US&NR=2024119261A1$$EView_record_in_European_Patent_Office$$FView_record_in_$$GEuropean_Patent_Office$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>Strudel, Robin</creatorcontrib><creatorcontrib>Grathwohl, Will S</creatorcontrib><creatorcontrib>Tallec, Corentin</creatorcontrib><creatorcontrib>Sifre, Laurent</creatorcontrib><creatorcontrib>Du, Yilin</creatorcontrib><creatorcontrib>Dieleman, Sander Etienne Lea</creatorcontrib><creatorcontrib>Savinov, Nikolay</creatorcontrib><creatorcontrib>Leblond, Rémi</creatorcontrib><creatorcontrib>Ganin, Iaroslav</creatorcontrib><creatorcontrib>Mensch, Arthur</creatorcontrib><creatorcontrib>Altché, Florent</creatorcontrib><title>DISCRETE TOKEN PROCESSING USING DIFFUSION MODELS</title><description>Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output sequence of discrete tokens using a diffusion model. In one aspect, a method includes generating, by using the diffusion model, a final latent representation of the sequence of discrete tokens that includes a determined value for each of a plurality of latent variables; applying a de-embedding matrix to the final latent representation of the output sequence of discrete tokens to generate a de-embedded final latent representation that includes, for each of the plurality of latent variables, a respective numeric score for each discrete token in a vocabulary of multiple discrete tokens; selecting, for each of the plurality of latent variables, a discrete token from among the multiple discrete tokens in the vocabulary that has a highest numeric score; and generating the output sequence of discrete tokens that includes the selected discrete tokens.</description><subject>CALCULATING</subject><subject>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</subject><subject>COMPUTING</subject><subject>COUNTING</subject><subject>PHYSICS</subject><fulltext>true</fulltext><rsrctype>patent</rsrctype><creationdate>2024</creationdate><recordtype>patent</recordtype><sourceid>EVB</sourceid><recordid>eNrjZDBw8Qx2DnINcVUI8fd29VMICPJ3dg0O9vRzVwgFky6ebm5Alr-fgq-_i6tPMA8Da1piTnEqL5TmZlB2cw1x9tBNLciPTy0uSExOzUstiQ8NNjIwMjE0tDQyM3Q0NCZOFQDegydl</recordid><startdate>20240411</startdate><enddate>20240411</enddate><creator>Strudel, Robin</creator><creator>Grathwohl, Will S</creator><creator>Tallec, Corentin</creator><creator>Sifre, Laurent</creator><creator>Du, Yilin</creator><creator>Dieleman, Sander Etienne Lea</creator><creator>Savinov, Nikolay</creator><creator>Leblond, Rémi</creator><creator>Ganin, Iaroslav</creator><creator>Mensch, Arthur</creator><creator>Altché, Florent</creator><scope>EVB</scope></search><sort><creationdate>20240411</creationdate><title>DISCRETE TOKEN PROCESSING USING DIFFUSION MODELS</title><author>Strudel, Robin ; Grathwohl, Will S ; Tallec, Corentin ; Sifre, Laurent ; Du, Yilin ; Dieleman, Sander Etienne Lea ; Savinov, Nikolay ; Leblond, Rémi ; Ganin, Iaroslav ; Mensch, Arthur ; Altché, Florent</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-epo_espacenet_US2024119261A13</frbrgroupid><rsrctype>patents</rsrctype><prefilter>patents</prefilter><language>eng</language><creationdate>2024</creationdate><topic>CALCULATING</topic><topic>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</topic><topic>COMPUTING</topic><topic>COUNTING</topic><topic>PHYSICS</topic><toplevel>online_resources</toplevel><creatorcontrib>Strudel, Robin</creatorcontrib><creatorcontrib>Grathwohl, Will S</creatorcontrib><creatorcontrib>Tallec, Corentin</creatorcontrib><creatorcontrib>Sifre, Laurent</creatorcontrib><creatorcontrib>Du, Yilin</creatorcontrib><creatorcontrib>Dieleman, Sander Etienne Lea</creatorcontrib><creatorcontrib>Savinov, Nikolay</creatorcontrib><creatorcontrib>Leblond, Rémi</creatorcontrib><creatorcontrib>Ganin, Iaroslav</creatorcontrib><creatorcontrib>Mensch, Arthur</creatorcontrib><creatorcontrib>Altché, Florent</creatorcontrib><collection>esp@cenet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Strudel, Robin</au><au>Grathwohl, Will S</au><au>Tallec, Corentin</au><au>Sifre, Laurent</au><au>Du, Yilin</au><au>Dieleman, Sander Etienne Lea</au><au>Savinov, Nikolay</au><au>Leblond, Rémi</au><au>Ganin, Iaroslav</au><au>Mensch, Arthur</au><au>Altché, Florent</au><format>patent</format><genre>patent</genre><ristype>GEN</ristype><title>DISCRETE TOKEN PROCESSING USING DIFFUSION MODELS</title><date>2024-04-11</date><risdate>2024</risdate><abstract>Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating an output sequence of discrete tokens using a diffusion model. In one aspect, a method includes generating, by using the diffusion model, a final latent representation of the sequence of discrete tokens that includes a determined value for each of a plurality of latent variables; applying a de-embedding matrix to the final latent representation of the output sequence of discrete tokens to generate a de-embedded final latent representation that includes, for each of the plurality of latent variables, a respective numeric score for each discrete token in a vocabulary of multiple discrete tokens; selecting, for each of the plurality of latent variables, a discrete token from among the multiple discrete tokens in the vocabulary that has a highest numeric score; and generating the output sequence of discrete tokens that includes the selected discrete tokens.</abstract><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | |
| ispartof | |
| issn | |
| language | eng |
| recordid | cdi_epo_espacenet_US2024119261A1 |
| source | esp@cenet |
| subjects | CALCULATING COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS COMPUTING COUNTING PHYSICS |
| title | DISCRETE TOKEN PROCESSING USING DIFFUSION MODELS |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-26T11%3A17%3A54IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-epo_EVB&rft_val_fmt=info:ofi/fmt:kev:mtx:patent&rft.genre=patent&rft.au=Strudel,%20Robin&rft.date=2024-04-11&rft_id=info:doi/&rft_dat=%3Cepo_EVB%3EUS2024119261A1%3C/epo_EVB%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |