Evaluation settings in Tables 2 and 3

[BunsenFeng]

Hi, Thank you for sharing the implementation of your wonderful work! It is a super valuable contribution to the research community.

I have a silly question about the "CNN/DM" and "XSUM" settings in Tables 2 and 3. Did you

a) train FactGraph on training examples in "All data" and evaluated test examples that belong to "CNN/DM" and "XSUM" separately
b) train FactGraph on training examples that belong to "CNN/DM" and "XSUM" and evaluated on test examples that belong to the corresponding source

We will make sure to cite your paper in our work. Looking forward to hearing from you!