>>106558636
Good evening maids and otherwise! A few threads back (>>106490115), Doctor Selig posted the weights and biases for a BBNN version of a half adder apparently discovered via brute-force Godel encoding. This struck my curiosity. Being one of those "undergrads who can't even understand what puremath cs is", I set out to recreate and confirm the result. Pic related shows the output of my hand-coded BBNN, loaded with Doctor Selig's parameters, for all valid inputs. The results speak for themselves. In that same thread, a full adder's Godel encoding was presented along with a neat graph of it's constituent nodes. Presumably, some function exists to translate this encoding into BBNN parameters. A hint from Doctor Selig regarding this translation would be appreciated. Regardless, with the internet and a copy of Wolfy MaidSearch in hand, my research continues.