SECTION / 03
Articles.
Articles and technical writing.
11 ENTRIES
01The Maths Behind Sora Video GenerationAn architecture-focused explanation of Sora that connects video generation to diffusion transformers, video vision transformers, variable-resolution inputs, and stochastic differential equations.02Top Deep Learning Papers of 2022A guided selection of notable 2022 deep learning papers, with explanations shaped by a computer vision and self-supervised learning perspective.03A Guide to Transformer ArchitecturesA practical guide to transformer building blocks, self-attention, and architecture variants across language and computer vision, with attention to how the pieces fit together.04Does AI Extrapolate? Or how a perfectly innocent machine learning question somehow ends with Gödel, God, and the informational substrate of reality05Give Me $120M and I’ll Change AIA long-form thought experiment about building an AI lab with a $120 million budget, set against the scale of the AI market and the cost of training models with limited adoption.06Why Do We Use MSE? Three Levels of ComplexityA three-level explanation of mean squared error: its role in regression, the gradient it produces, and what its curvature tells us about optimization.07How to Make Your LLM ReasonA written companion to Diego’s video series that connects language-model reasoning with reinforcement learning and modern deep learning methods.08Tokenization in NLP: An Academic GuideAn introduction to how language models turn raw text into tokens and numerical representations, including word and subword units and Word2Vec-style embeddings.09Deep Learning Advanced Math Simplified: A Self-Taught Guide to Topology and High-Dimensional GeometryA self-study guide to topology and high-dimensional geometry, building intuition for manifolds and why these ideas help frame deep learning representations.10A Guide into Activation Functions in Neural NetworksAn accessible guide to how activation functions add nonlinearity to neural networks, shape their outputs, and affect gradient flow during training.112023 and Beyond: The Latest Trends and Advances in Computer VisionPart one of a guided survey of computer vision advances discussed in 2023, including visual language models and the growing overlap between vision and language.