mantasu/cs231n
Student solutions for Stanford CS231n course assignments covering convolutional neural networks, image captioning with RNNs and Transformers, self-supervised learning, and Denoising Diffusion Probabilistic Models.

This repository provides concise solutions to Stanford’s CS231n visual recognition course assignments from 2021-2025. Assignments cover fundamental deep learning topics including k-NN classifiers, softmax, two-layer networks, convolutional architectures, BatchNorm, Dropout, and PyTorch implementations on CIFAR-10. More advanced assignments include image captioning with Vanilla RNNs and Transformers, self-supervised learning for image classification, and Denoising Diffusion Probabilistic Models (DDPM).