If they don't actually help then the attention weight for that connection would ... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		eachro on Sept 19, 2023 \| parent \| context \| favorite \| on: Graph Neural Networks use graphs when they shouldn... If they don't actually help then the attention weight for that connection would tend to 0, right? Then it becomes a problem of overfitting which we have a large arsenal to combat.

SpaceManNabs on Sept 19, 2023 [–]

> If they don't actually help then the attention weight for that connection would tend to 0, right?

Not necessarily. That is what this paper is about. From what I understand, they also consider graph attention networks.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact