Add a "being attacked" input neuron

[ZRunner]

Once #8 has been implemented, add an input neuron telling if the subject has recently been attacked. The value should be 0.0 if nothing happened recently (recently being up to you to decide), 1.0 if an attack just happened.