Kiruko

Kiruko offers in-depth instruction on how computational processes function, from foundational elements to advanced architectures. Key learning areas include: * Tokenization techniques like BPE * Self-attention mechanisms (Query/Key/Value) * Transformer architecture comprehension * Understanding context windows and generative model behaviors * Reinforcement learning from human feedback (RLHF) * Retrieval-augmented generation (RAG) The app presents these concepts through interactive lessons designed for clarity. Visual explanations, engaging mini-games, and graded exams reinforce understanding, including unique "speak-it-out" and interview sections to practice explaining complex ideas in your own words. The entire experience is tailored for deep comprehension, moving beyond surface-level explanations. Built for anyone seeking a comprehensive understanding of advanced computational models without requiring coding knowledge. Whether you are a student, professional, or simply curious about the internal workings of these systems, Kiruko provides a structured and accessible learning path. Master the core principles behind modern generative systems through a guided, engaging interface that makes complex topics approachable.