You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello, I would like to inquire about how exactly the activation is adjusted during the inference process by multiplying the weights by a number and dividing the activations by a number. As far as I know, it is only possible to obtain the output of a particular layer, but not to adjust the activation of a specific input
The text was updated successfully, but these errors were encountered:
Hello, I would like to inquire about how exactly the activation is adjusted during the inference process by multiplying the weights by a number and dividing the activations by a number. As far as I know, it is only possible to obtain the output of a particular layer, but not to adjust the activation of a specific input
The text was updated successfully, but these errors were encountered: