4 ms·
Can't believe that FNet paper flew under my radar all this time—what a cool idea! It's remarkable to me that it works so well considering I haven't heard anyone
by mkaic 2y ago
Can't believe that FNet paper flew under my radar all this time—what a cool idea! It's remarkable to me that it works so well considering I haven't heard anyone mention it before! Do you know if any follow-up work was done?
- CGamesPlay 2y agoThis is the paper, which appears to be cited in hundreds of others, some of which appear to be about efficiency gains. https://arxiv.org/abs/2105.03824 https://arxiv.org/abs/2105.03824
- westurner 2y ago"Fnet: Mixing tokens with fourier transforms" (2021) https://arxiv.org/abs/2105.03824 https://arxiv.org/abs/2105.03824 https://scholar.google.com/scholar?cites=14236996275885084866&as_sdt=5,43&sciodt=0,43&hl=en https://scholar.google.com/scholar?cites=1423699627588508486... Fourier Transform and convolution. Shouldn't a deconvolvable NN be more explainable? #XAI Deconvolution: https://en.wikipedia.org/wiki/Deconvolution https://en.wikipedia.org/wiki/Deconvolution
- pests 2y agoI know AI is moving fast but they were only published within the last month or three.