Academic Journal
Προβολή σε EDS

Rose: Composable Autodiff for the Interactive Web (Artifact)

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Rose: Composable Autodiff for the Interactive Web (Artifact)
Συγγραφείς: Estep, Sam, Ni, Wode, Rothkopf, Raven, Sunshine, Joshua
Συνεισφορές: Sam Estep and Wode Ni and Raven Rothkopf and Joshua Sunshine
Στοιχεία εκδότη: Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Έτος έκδοσης: 2024
Συλλογή: DROPS - Dagstuhl Research Online Publication Server (Schloss Dagstuhl - Leibniz Center for Informatics )
Θεματικοί όροι: Automatic differentiation, differentiable programming, compilers, web
Περιγραφή: Reverse-mode automatic differentiation (autodiff) has been popularized by deep learning, but its ability to compute gradients is also valuable for interactive use cases such as bidirectional computer-aided design, embedded physics simulations, visualizing causal inference, and more. Unfortunately, the web is ill-served by existing autodiff frameworks, which use autodiff strategies that perform poorly on dynamic scalar programs, and pull in heavy dependencies that would result in unacceptable webpage sizes. This document accompanies the research paper introducing Rose, a lightweight autodiff framework for the web using a new hybrid approach to reverse-mode autodiff, blending conventional tracing and transformation techniques in a way that uses the host language for metaprogramming while also allowing the programmer to explicitly define reusable functions that comprise a larger differentiable computation. We demonstrate the value of the Rose design by porting two differentiable physics simulations, and evaluate its performance on an optimization-based diagramming application, showing Rose outperforming the state-of-the-art in web-based autodiff by multiple orders of magnitude.
Τύπος εγγράφου: article in journal/newspaper
Περιγραφή αρχείου: application/pdf
Γλώσσα: English
Relation: Is Part Of DARTS, Volume 10, Issue 2, Special Issue of the 38th European Conference on Object-Oriented Programming (ECOOP 2024); https://drops.dagstuhl.de/entities/document/10.4230/DARTS.10.2.7
DOI: 10.4230/DARTS.10.2.7
Διαθεσιμότητα: https://doi.org/10.4230/DARTS.10.2.7
https://nbn-resolving.org/urn:nbn:de:0030-drops-209053
https://drops.dagstuhl.de/entities/document/10.4230/DARTS.10.2.7
Rights: https://creativecommons.org/licenses/by/4.0/legalcode
Αριθμός Καταχώρησης: edsbas.6619797C
Βάση Δεδομένων: BASE
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://doi.org/10.4230/DARTS.10.2.7#
    Name: EDS - BASE (ns324271)
    Category: fullText
    Text: View record from BASE
Header DbId: edsbas
DbLabel: BASE
An: edsbas.6619797C
RelevancyScore: 945
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 944.983581542969
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Rose: Composable Autodiff for the Interactive Web (Artifact)
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Estep%2C+Sam%22">Estep, Sam</searchLink><br /><searchLink fieldCode="AR" term="%22Ni%2C+Wode%22">Ni, Wode</searchLink><br /><searchLink fieldCode="AR" term="%22Rothkopf%2C+Raven%22">Rothkopf, Raven</searchLink><br /><searchLink fieldCode="AR" term="%22Sunshine%2C+Joshua%22">Sunshine, Joshua</searchLink>
– Name: Author
  Label: Contributors
  Group: Au
  Data: Sam Estep and Wode Ni and Raven Rothkopf and Joshua Sunshine
– Name: Publisher
  Label: Publisher Information
  Group: PubInfo
  Data: Schloss Dagstuhl – Leibniz-Zentrum für Informatik
– Name: DatePubCY
  Label: Publication Year
  Group: Date
  Data: 2024
– Name: Subset
  Label: Collection
  Group: HoldingsInfo
  Data: DROPS - Dagstuhl Research Online Publication Server (Schloss Dagstuhl - Leibniz Center for Informatics )
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Automatic+differentiation%22">Automatic differentiation</searchLink><br /><searchLink fieldCode="DE" term="%22differentiable+programming%22">differentiable programming</searchLink><br /><searchLink fieldCode="DE" term="%22compilers%22">compilers</searchLink><br /><searchLink fieldCode="DE" term="%22web%22">web</searchLink>
– Name: Abstract
  Label: Description
  Group: Ab
  Data: Reverse-mode automatic differentiation (autodiff) has been popularized by deep learning, but its ability to compute gradients is also valuable for interactive use cases such as bidirectional computer-aided design, embedded physics simulations, visualizing causal inference, and more. Unfortunately, the web is ill-served by existing autodiff frameworks, which use autodiff strategies that perform poorly on dynamic scalar programs, and pull in heavy dependencies that would result in unacceptable webpage sizes. This document accompanies the research paper introducing Rose, a lightweight autodiff framework for the web using a new hybrid approach to reverse-mode autodiff, blending conventional tracing and transformation techniques in a way that uses the host language for metaprogramming while also allowing the programmer to explicitly define reusable functions that comprise a larger differentiable computation. We demonstrate the value of the Rose design by porting two differentiable physics simulations, and evaluate its performance on an optimization-based diagramming application, showing Rose outperforming the state-of-the-art in web-based autodiff by multiple orders of magnitude.
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: article in journal/newspaper
– Name: Format
  Label: File Description
  Group: SrcInfo
  Data: application/pdf
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: NoteTitleSource
  Label: Relation
  Group: SrcInfo
  Data: Is Part Of DARTS, Volume 10, Issue 2, Special Issue of the 38th European Conference on Object-Oriented Programming (ECOOP 2024); https://drops.dagstuhl.de/entities/document/10.4230/DARTS.10.2.7
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.4230/DARTS.10.2.7
– Name: URL
  Label: Availability
  Group: URL
  Data: https://doi.org/10.4230/DARTS.10.2.7<br />https://nbn-resolving.org/urn:nbn:de:0030-drops-209053<br />https://drops.dagstuhl.de/entities/document/10.4230/DARTS.10.2.7
– Name: Copyright
  Label: Rights
  Group: Cpyrght
  Data: https://creativecommons.org/licenses/by/4.0/legalcode
– Name: AN
  Label: Accession Number
  Group: ID
  Data: edsbas.6619797C
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsbas&AN=edsbas.6619797C
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.4230/DARTS.10.2.7
    Languages:
      – Text: English
    Subjects:
      – SubjectFull: Automatic differentiation
        Type: general
      – SubjectFull: differentiable programming
        Type: general
      – SubjectFull: compilers
        Type: general
      – SubjectFull: web
        Type: general
    Titles:
      – TitleFull: Rose: Composable Autodiff for the Interactive Web (Artifact)
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Estep, Sam
      – PersonEntity:
          Name:
            NameFull: Ni, Wode
      – PersonEntity:
          Name:
            NameFull: Rothkopf, Raven
      – PersonEntity:
          Name:
            NameFull: Sunshine, Joshua
      – PersonEntity:
          Name:
            NameFull: Sam Estep and Wode Ni and Raven Rothkopf and Joshua Sunshine
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2024
          Identifiers:
            – Type: issn-locals
              Value: edsbas
            – Type: issn-locals
              Value: edsbas.oa
ResultId 1