There are bunches of types of latches and flip-flops, and several of them can be implemented with the Threshold Logic Neuron. The common thread is the technique of feeding one of the (identical) axon outputs back into one of the dendrites with a patch cable. Depending on that dendrite’s weight, it can either force the sum past the threshold and latch the output, or it can force the sum lower by itself, or in conjunction with another dendrite that can act as a “reset” input. The wiki has a brief “cookbook” of Threshold Logic settings, which gets into that a little bit, and hopefully gets you thinking about other possibilities. 
Oh, and also, the Neural Net full app has a data flip-flop, a toggle flip-flop, and a latch built in, so check that out if you love logic!