🔮 Deep Learning & Neural Architectures: Sub-Curriculum Index

Welcome to the Deep Learning & Neural Architectures curriculum. This track covers the foundational architectures that power modern artificial intelligence: backpropagation, deep feedforward nets, convolutional neural networks (CNNs) in computer vision, sequence models in NLP, Reinforcement Learning with Human Feedback (RLHF), generative models (VAEs, Diffusion), Transformer self-attention, and vision-language multimodal models.

Every guide in this series strictly follows a two-part learning format:

  • ⚡ Quick Dive: Architecture diagrams, activation functions, tensor shape transformations, and PyTorch one-liners.
  • 📖 Extended Guide: Mathematical derivations, backprop computational graphs, attention mechanisms, and PyTorch model implementations.

📚 Curriculum Roadmap

# Guide Primary Topics Covered
01 Deep Learning Fundamentals Universal approximation theorem, deep vs. shallow networks, vanishing/exploding gradients, and normalization layers.
02 Neural Networks & Backpropagation Computational graphs, chain rule automatic differentiation, loss functions, and optimizers (SGD, Adam, AdamW).
03 NLP Foundations & Word Vectors Tokenization (BPE, WordPiece), Word2Vec, GloVe, Recurrent Neural Networks (RNNs), and LSTMs.
04 Computer Vision & CNNs Spatial convolution kernels, pooling layers, ResNet skip connections, object detection (YOLO), and Vision Transformers (ViT).
05 Reinforcement Learning & RLHF Markov Decision Processes (MDP), Q-Learning, Policy Gradients (PPO), Reward Modeling, and RLHF for LLM alignment.
06 Generative Models: VAEs & Diffusion Latent space representations, Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), and Denoising Diffusion (DDPM).
07 Transformers & Attention Mechanisms Scaled Dot-Product Attention, Multi-Head Attention (MHA), Rotary Position Embeddings (RoPE), and FlashAttention.
08 Multimodal Architectures Contrastive Language-Image Pretraining (CLIP), cross-attention fusion, image projection layers, and vision-language LLMs (LLaVA).
⌂ All Guides