I’m taking CS231n 2017 and doing the HW assignments from cs231n 2024. In every Q, specifically SVM and Softmax , I can easily calculate loss but always find it difficult to calculate the gradient whether wrt to W or X. I think I’m missing something but I’m not sure. I was hoping if someone could walk me through the steps step by step. I am very desperate 🙁

PS: I know matrices, calculus and etc.

In previous problems, I tried taking derivative of the lost functions and then implement it in code. However, for softmax and svm, I seem to face difficulty in implementing it. I was hoping someone could walk me through the mathematical steps of how to go about it.