There are different coordinate spaces for this but as the documentation says: "This is the vector from the vertex or fragment to the camera". So what has the coordinate space - Object, View, World, Tangent, got to do with it - what does each represent? Is it not just from the current world-space pixel on object to the camera. ideally in view-space after objects transformed around camera? Also it is called 'direction' which you would assume means normalized? But it says vector 'from' which implies ObjectFromCameraVector not ViewDirection? Also I never know if shader graphs are all per-pixel 'fragment-shader' based?