dkhonker

> I'm not sure if I understand the question correctly. But I think you do not have to use a dense adjacency matrix as input. Node features themselves are enough...

“ We would get the same results if we're solving via OLS ” Is there any code for this part?

I have the same doubt. I tried simply adding a linear layer to the output, but the result was even worse than not using it. 我同样有这个疑惑，我尝试在output简单地添加一个线性层，然而效果甚至不如不使用。