% IMPORTANT: The following is UTF-8 encoded.  This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.

@INPROCEEDINGS{Lohoff:1038128,
      author       = {Lohoff, Jamie and Kaya, Anil and Assmuth, Florian and
                      Neftci, Emre},
      title        = {{A} {T}ruly {S}parse and {G}eneral {I}mplementation
                      of{G}radient-{B}ased {S}ynaptic {P}lasticity},
      reportid     = {FZJ-2025-01175},
      pages        = {n/a},
      year         = {2025},
      note         = {Accepted as an oral presentation.},
      abstract     = {Online synaptic plasticity rules derived from gradi-ent
                      descent achieve high accuracy on a wide range of
                      practicaltasks. However, their software implementation often
                      requirestediously hand-derived gradients or using gradient
                      backprop-agation which sacrifices the online capability of
                      the rules. Inthis work, we present a custom automatic
                      differentiation (AD)pipeline for sparse and online
                      implementation of gradient-based synaptic plasticity rules
                      that generalizes to arbitraryneuron models. Our work
                      combines the programming easeof backpropagation-type methods
                      for forward AD while beingmemory-efficient. To achieve this,
                      we exploit the advantageouscompute and memory scaling of
                      online synaptic plasticity byproviding an inherently sparse
                      implementation of AD whereexpensive tensor contractions are
                      replaced with simple element-wise multiplications if the
                      tensors are diagonal. Gradient-basedsynaptic plasticity
                      rules such as eligibility propagation (e-prop)have exactly
                      this property and thus profit immensely from thisfeature. We
                      demonstrate the alignment of our gradients withrespect to
                      gradient backpropagation on an synthetic task wheree-prop
                      gradients are exact, as well as audio speech
                      classificationbenchmarks. We demonstrate how memory
                      utilization scales withnetwork size without dependence on
                      the sequence length, asexpected from forward AD methods.},
      month         = {Mar},
      date          = {2025-03-25},
      organization  = {Neuro-Inspired Computational Elements,
                       Heidelberg (Germany), 25 Mar 2025 - 28
                       Mar 2025},
      cin          = {PGI-15},
      cid          = {I:(DE-Juel1)PGI-15-20210701},
      pnm          = {5234 - Emerging NC Architectures (POF4-523) / BMBF 16ME0400
                      - Verbundprojekt: Neuro-inspirierte Technologien der
                      künstlichen Intelligenz für die Elektronik der Zukunft -
                      NEUROTEC II - (16ME0400) / GREENEDGE - Taming the
                      environmental impact of mobile networks through GREEN EDGE
                      computing platforms (953775)},
      pid          = {G:(DE-HGF)POF4-5234 / G:(BMBF)16ME0400 /
                      G:(EU-Grant)953775},
      typ          = {PUB:(DE-HGF)8},
      doi          = {10.34734/FZJ-2025-01175},
      url          = {https://juser.fz-juelich.de/record/1038128},
}