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
Thanks for the greatwork! It is creative and the shown results are promising.
I saw from paper you have several baselines to be compared against your proposed CSMN (e.g. 1-nearest neighbor to user contents, RNN seq2seq with active vocabulary)
Would you release the implementations of those baselines?
Besides that, given the recent advancement on NLP (transformer, GPT-2 ... etc) , would you (and how would you) propose your CSMN differently under modern context (as of 2020)?
The text was updated successfully, but these errors were encountered:
Thanks for the greatwork! It is creative and the shown results are promising.
I saw from paper you have several baselines to be compared against your proposed CSMN (e.g. 1-nearest neighbor to user contents, RNN seq2seq with active vocabulary)
Would you release the implementations of those baselines?
Besides that, given the recent advancement on NLP (transformer, GPT-2 ... etc) , would you (and how would you) propose your CSMN differently under modern context (as of 2020)?
The text was updated successfully, but these errors were encountered: