Where to go next

• torch.einsum / tf.einsum / jnp.einsum — identical syntax to what you used in section 06.
• np.linalg — the rest of Chapter 2: eigendecomposition, lstsq, pinv, qr, cholesky.
• scipy.signal and skimage.restoration — convolution and deconvolution beyond today.
• tensorly — proper implementations of Tucker and CP for real projects.
• The rest of the four deep dives you opened live: 13 (convolution and deconvolution), 14 (CP), 17 (attention) and 20 (AlphaTensor). Each goes well past the route the room worked.
• An open problem at the end of section 11: a 9 × 9 × 9 tensor of 27 ones whose rank nobody knows, and why a cheaper matrix multiplication is worth finding.
• References and further reading — the linear algebra books, the Kolda & Bader survey, and the Tucker, CP/PARAFAC and Eckart–Young papers.

The machine-generated companion is a different kind of thing: it is generated, not written. Check it against the handbook rather than trusting it over it.