Back propagation neural network

I learnt that the output of activation functions, logsig and tansig, returns the value between [0 1] and [-1 1] respectively. What will happen if the target values are beyond these limits?

[Mohammad Sami]

One of the reasons is that larger values can result in a problem of exploding gradients, when training the network.

[Sivamani S]

Dear Mohammed Sami,

I gave 3 inputs and 1 target value in BPNN. I obtained values of weights and bias. When I fit manually weights and bias to the ANN model, I am not getting output what ANN provides. I tried for 1 sample run. Is there any way to check the ANN output manually?