Implementation doubt

[helia95]

Hi,

I quite don't understand where Eq. (4), (7) of the paper come in this implementation.
Can you point me to it?

Thanks

[shunk031]

Hi, @helia95 Thank you for the comment.

Is the Eq. (4, 7) that you are pointing out actually Eq. (7) in the paper?
I'm writing about the Eq. (7) in the paper below.

I'm really sorry, but there is an implementation mistake in the current main branch (and the latest commit) and there is no corresponding implementation for the Eq. (7).

I'm currently implementing the equation as #8 (and the specific lines are here). Thank you for your interest.

[ethansmith2000]

@shunk031 your implementation looks correct there, but the one thing i am still unsure about is that it looks like you've modified both the keys and values with their method, but the graphic on the paper and the equation makes it look like only the values are modified, not the keys.

in the sum, M_t is constant and dependent on step, but we iterate through all the newly modified values with their method.

However, i have not tried your pipeline out of the box, do you have any results to share?