How the malmul works in this library?

[pigu163]

Hi, I am using pytorch_sparse matmul, but see abnormal result

from torch_sparse import SparseTensor, matmul

x [batch, features, nodes, step]

x = torch.rand(100, 20, 3, 24)
adj = torch.eye(100,100)
adj = SparseTensor.from_dense(adj)
out = adj.matmul(x) --> torch.Size([100, 20, 100, 24])

@rusty1s
Generally, sparse matrix (100, 100), it can't directly multiply with the (3, 24) dimensions because the dimensions don't align properly for matrix multiplication, but there is no error pop up. If this operator same as torch.matmul, and also follow their broadcast rules, it shouldn't be like this.
Is there any explanation for this operator ?

Beside, torch.matmut has bug on broadcast sparse.matrix times dense.matrix since torch verison >= 2.0.1

[github-actions]

This issue had no activity for 6 months. It will be closed in 2 weeks unless there is some new activity. Is this issue already resolved?