Twitter | Pretraživanje | |
Gary Marcus 30. stu
thread on universally-quantified one to one mappings, which I still see as key bottleneck to future progress.
Reply Retweet Označi sa "sviđa mi se"
Gary Marcus 30. stu
Odgovor korisniku/ci @AdamMarblestone
I am not sure it is a general solution but building in operations over variables as NTM's do is the crux of what I have been arguing for. IIRC i mentioned that specific work approvingly in my Deep Learning Critical Appraisal and follow up on Medium (In Defense of Skepticism).
Reply Retweet Označi sa "sviđa mi se"
Adam Marblestone 10. pro
Odgovor korisniku/ci @GaryMarcus
Yes, and FYI I think transformers with “relative position encoding” also arguably build in a syntactic/content independent variable binding mechanism
Reply Retweet Označi sa "sviđa mi se"
Gary Marcus 30. stu
Odgovor korisniku/ci @AdamMarblestone
to my mind the question has never been is the brain a neural network but rather what kind, and I have always argued that it must be a neural network that had operations over variables. even when I co-authored with you :)
Reply Retweet Označi sa "sviđa mi se"