undefined | Better HN

0 pointstaneq4y ago0 comments

Can we barely model it yet? I thought they weren’t at the point of getting able to simulate any actual work behaviours?

0 comments

3 comments · 1 top-level

katmannthree4y ago· 2 in thread

We can model neural activations but the whole-cell body simulation was still unstable last I checked (admittedly a year or so ago)

taneqOP4y ago

Does the modeled activity track any real world measurements, though? If we're going to claim any real veracity from a simulation then we need to crosscheck it against real measurements, right?

katmannthree4y ago

I believe they’ve attempted to measure synapse weights from real worms for training the net, the firings themselves are just basic electrochemistry.

1 more reply

j / k navigate · click thread line to collapse