I understand your question. Sometimes folks can answer tersely when they don't clearly read the question, but we like to have fun here. You are using the equation from Chapter 8 of Garber & Hoel:
Gp = 3.2 + L/Sp + [2.7 *(Nped/We)] for We > 10 ft
or
Gp = 3.2+ L/Sp + (0.27Nped) for We <=10 ft
So, you clearly understand that the Sp factor here is going to be either 3.5 fps or 4 fps, the L is obviously the distance across the road, and the Nped is what you are trying make sure you understand.
This is actually a very valid question, would you assume the bicyclists to be Class B and C cyclists and treat them as pedestrians when crossing the intersection (i.e., they get off their bike and walk it across the intersection looking both ways, like most people were taught when they were kids and many people still do who aren't comfortable riding their bikes in traffic). If you do that, then you would theoretically increase the Nped value to be number of true pedestrians plus number of bicyclists that you assume would dismount and walk. Or, do you consider the bicyclists to be class A cyclists, who consider bikes as vehicles and ride in the vehicular travel way, feel comfortable riding with traffic, in which case you would not count the bikes with the peds.
I think in reality you would be a bit more conservative and treat the bicyclists as peds for the crossing, but it seems that in this case they expect you to use the pedestrian value as it is given in the problem.