AI Associate Exam Questions and Answers 100% Pass

AI Associate Exam Questions and Answers 100% Pass Transformer - AI Architecture that is capable of identifying important relationships between words, even if those words are far apart in a block of text. Retains connection even after processing a lot of words - Create outputs based on sequential data rather than data points.. E.g. paragraphs. Parallel Computing - Multiple computer processors doing calculations at once - one doing the first calculation, one doing the second, etc. Basically just makes it faster/adds computational power. GPT - Generative pre-trained transformer Tech stack "layers" making generative AI possible - - Computer hardware providers: providing the computing power. 2 Katelyn Whitman, All Rights Reserved © 2025 - Cloud Platforms: allow developers to tap into compute hardware in a cloud deployment model... can rent the hardware - AI Models (including LLMs) - developers can connect to LLMs through APIs. The trained and accessible AI model is commonly referred to as a foundational model. Devs can switch from one foundational model to another (b/c it's via an API) - Infrastructure optimization: providing tools/services that make for efficient, higher quality model training. E.g., curated datasets; analytics. - At the tip of the iceberg... the applications themselves Generative Adversarial Networks (GANs) - Made of two neural networks- a generator and a discriminator. Generator creates; discriminator tries to determine if it's real or fake. Work with each other to create realistic content Natural Language Processing - contains two subfields: - NLU: Natural Language Understanding (involves processing unstructured data to structured, to interpret meaning) 3 Katelyn Whitman, All Rights Reserved © 2025 - NLG: Natural Language Generation (involves processing structured data to unstructured - to generate human-like language) Parsing - Techniques to process and analyze text are called parsing. This involves breaking down text or speech into smaller parts. Syntactic parsing: elements of NL analyzed to identifying grammatical structure; Semantic parsing: derives meaning Syntactic parsing might include - Segmentation: divide into chunks Tokenization: sentences split into individual words. Simpler in english b/c of spaces... Stemming: reducing words to root form. Sometimes inaccurate due to words w/ same spelling Lemmetization: like stemming but uses speech to arrive at