See tensors.
Think in ML.
The study companion to a six-hour, hands-on workshop on reshaping, solving, convolving and factorizing tensors, in NumPy on real images, video and city data. Work through it unit by unit, then test yourself.
From the workshop by Ravi Kalia and Sebastian Laverde Chunza · 22 study units, from the linear algebra you need to AlphaTensor
What you'll study
Before the workshop
Check the linear algebra the workshop builds on, then set up your runtime and meet the nine real datasets.
Parts I–III · Reading tensors
What a tensor is, what every axis counts, and how to index, broadcast, reshape and transpose real data without silent bugs.
Part IV · Contract, solve, factorize
einsum, the pseudoinverse, recursion, matrix factorizations, Tucker and the tensor factorizations, then the idea that connects them.
Deep dives
The live labs and take-home deep dives: convolution, CP by hand, generalized CP, PCA from tensors, multi-head attention, feature compression, CP or Tucker in practice, and AlphaTensor.
Ready when you are.
Free to start studying. No account needed to begin your first deck.
Start studying - free