Vorticey - have you been drinking? pushing the collective forward? airflow throught the disc offset? que?
Kota - when you flare the aircraft by pulling the cyclic back you reduce the induced flow by changing the angle at which the air meets the blades (relative airflow). This increases the angle of attack and thus the lift across the disc - the blades want to cone up due to the extra lift produced and as they do so the C of G of the blades moves inwards and conservation of angular momentum (the skater analogy) increases the RPM as a result.
The rate of change of the angle of attack determines how much pitch up and increase in rotor thrust (and NR) you get and the faster you are going the more effective it is (ie at low forward speeds a larger cyclic movement is required to produce the same increase compared to high forward speeds where a small aft cyclic input will produce large changes).
When you push the cyclic forward in auto, the reverse is exactly true - the angle of the rate of descent flow is reduced(inflow angle), therefore the AoA of the blades is reduced, they 'uncone' (for want of a better phrase - flap down is probably better), the C of G moves outwards and the blades slow down. As with flare effect the more aggresive the cyclic input the greater the change.
Once you have made the cyclic input and stabilised the attitude, eventually the Nr will recover - you have increased you speed in auto and will experience an increase in RoD which will gradually increase the inflow angle again, restoring the AoA.