dl-m9/llm180

LLM180 is a study tool built for people preparing for machine learning and large language model job interviews, especially internship interviews at Chinese AI companies. It packages interview preparation into a single web page you can open directly in your browser, with no backend or build step required. The page contains two separate question sets: 180 conceptual interview questions covering large language model theory, and 40 coding style questions where you write out an implementation by hand. The conceptual questions are designed to be answered out loud in about 60 to 90 seconds, and each one comes with a full answer explaining the underlying mechanism, how it is actually used in real systems, and the risks involved in evaluating it. The coding questions include the problem, a suggested approach, a reference solution in Python or SQL, complexity analysis, and common mistakes. The interface works like a slide deck. Only one question shows at a time, and you flip through with arrow keys or on screen buttons. You can switch between the two question sets from the top navigation, and reference code stays hidden until you have tried to answer yourself. Your progress, including which questions you marked as not understood, unclear, or mastered, is saved locally in your browser so it persists across sessions. Topics covered include reinforcement learning methods used to fine tune language models, retrieval augmented generation and AI agents, inference systems like KV caching and attention optimizations, training infrastructure such as distributed training frameworks, and multimodal models that combine vision and language. There is also a section aimed at specific companies' hiring processes. The author states this is a training tool built from real interview experiences but cross checked against papers and official documentation, not a research survey, so the standard answers prioritize accuracy over trends.