question about "visual sentinel"

[wzn0828]

Dear Jiasen Lu,
Thank you for your work on "Knowing When to Look: Adaptive Attention via A Visual Sentinel for Image Captioning".
I am writing to ask about the "visual sentinel": what is the difference between your "visual sentinel" and the hidden state ht？
I think your visual sentinel "st" and the LSTM's hidden state "ht" are the same except for the different symbols. Am I right ? If I am wrong, could you kindly give me some further explanation?

Many thanks in advance for your answer.
Kind regards
Zhennan Wang

[jamiechoi1995]

@wzn0828
In my opinion,
the formulations of "ht" and "st" are similar,
but they are affected by different variables when backpropagating loss,
which results in their different effects.

[jiasenlu]

Yes, I agree with @jamiechoi1995 Base on the different model inductive bias you impose, the weight will learn different functions.

[wzn0828]

@jamiechoi1995 @jiasenlu Thank you all, I think I have gotten a deeper understanding of this sentinel with your help.