The problem with neuroscience is not that we don’t know the answers, it is that we don’t yet even know the right questions. Speaking As a formally trained computer scientist who had to ingest neuroscience full time for 4+ years.
this pointer inadvertently highlights how challenging it is to connect ANN’s to biology. Haxby proposal here relates short term memory to voxels (each of which contains 50k+ neurons), whereas eg LSTMs treat memories as neuron level constructs. no decisive evidence for eitherhttps://twitter.com/jsalsman/status/1200278581118586880 …
-
-
- 3 more replies
New conversation -
-
-
Because none of the NN models are based on real neuronal anatomy. The brain is not just nodes & wires. No neurons actually touch. If they did, you’d have constitutive activation. Also, there is no accounting for secondary & tertiary hormonal signaling. Memories are fragmented.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.