Graph Neural Networks Quiz

Answers and explanations appear as you take the quiz.

1. What is the core computational mechanism of most graph neural networks?
   • Message passing and aggregation between neighboring nodes
   • Sliding a convolution over a fixed pixel grid
   • Autoregressive next-token prediction
   • Random walks with no learning
2. Why must a GNN's neighbor-aggregation function be permutation invariant?
   • To make training faster
   • Because a graph's neighbors have no inherent ordering
   • Because graphs are always directed
   • To reduce the number of parameters to one
3. What does 'over-smoothing' refer to in deep GNNs?
   • Images becoming blurry
   • The graph losing its edges
   • Node embeddings converging to be nearly identical as layers increase
   • Gradients exploding to infinity
4. How does a Graph Attention Network differ from a basic GCN layer?
   • It removes all graph edges before message passing
   • It assigns each node a unique text token
   • It guarantees deeper networks cannot over-smooth
   • It learns attention coefficients for neighbors, while the GCN uses degree-normalized propagation weights
5. In a GNN applied to a molecule, what do nodes and edges typically represent?
   • Atoms and chemical bonds
   • Pixels and colors
   • Words and sentences
   • Users and posts
6. How many hops away can information travel to a node after L message-passing layers?
   • Always exactly one hop
   • Up to L hops
   • The entire graph in a single layer
   • Zero hops
7. What makes GraphSAGE an inductive graph-learning method?
   • It stores a separate free parameter vector for every possible test node
   • It requires one fixed graph and cannot process new nodes
   • It learns neighborhood aggregation functions that can be applied to unseen nodes
   • It replaces node features with graph-level labels
8. In the original GCN propagation rule, what does symmetric degree normalization do to adjacency messages?
   • Sorts neighboring nodes alphabetically
   • Deletes every edge with a low feature value
   • Makes the graph directed by reversing all edges
   • Scales each edge contribution using the inverse square roots of the source and target node degrees