MachineLearning From Derivatives to Adam: Building Optimization Intuition with PyTorch From first and second derivatives to Adam optimizer: a hands-on walkthrough of key optimization ideas using PyTorch. Simple, minimal examples with gradients, momentum, and adaptive learning strategies to build deep intuition.
GenerativeAI Modeling the World with Tensors and Neural Networks We walk through the evolution from simple feedforward networks to powerful transformers and graph models, showing how richer data structures demand more sophisticated architectures. More importantly, we uncover how embedding spaces allow different modalities to interact.
