Farzad KaramiIntroduction to the Structure and Mathematics of Text-to-Image ModelsThis blog post looks into various text-to-image models, emphasizing their structural and mathematical foundations. These models represent a…Jan 42Jan 42
Farzad KaramiIntroducing Glaze: A Revolutionary Method to Protect Artists from Style Mimicry by Text-to-Image…Presented at the USENIX Security Symposium in 2023, the paper “Glaze: Protecting Artists from Style Mimicry by Text-to-Image Models” has…Aug 22, 2023Aug 22, 2023
Farzad KaramiStarlight, Traffic Jams, and The Shocking Power of Perspective: A Personal OdysseyThere are instances in life, moments of revelation, akin to those ‘aha’ moments when a truth or concept suddenly becomes apparent. These…Jul 21, 2023Jul 21, 2023
Farzad KaramiUnderstanding Graph Attention Networks: A Practical ExplorationDuring my recent experimentation with Graph Convolutional Networks (GCNs) and Graph Attention Networks (GATs), I observed an interesting…Jul 20, 20231Jul 20, 20231
Farzad KaramiHow Is ChatGPT’s Behavior Changing over Time?Large Language Models (LLMs) such as GPT-3.5 and GPT-4 have become increasingly prevalent in various applications, ranging from natural…Jul 20, 2023Jul 20, 2023
Farzad KaramiDecoding the Magic of Self-Attention: A Deep Dive into its Intuition and MechanismsThe self-attention mechanism is a key component in modern machine learning models, particularly when dealing with sequential data. This…Jul 10, 20231Jul 10, 20231