Question

What will happen if we initialize all the weights to 0 in neural Networks?
check all that apply
A. Model will not be able to learn
B. model will learn faster
C. model will fall into symmetric gradient condition
D.All the weights will be updated by different values .​

Answer 1

Answer:

Initializing all the weights with zeros leads the neurons to learn the same features during training. ... Thus, both neurons will evolve symmetrically throughout training, effectively preventing different neurons from learning different things

Explanation:

pls mark me as brainlst

Answer 2

Answer:

Explanation:

refer to the site deeplearning

ai-notes/initialization/