Algorithm: In machine learning, algorithms refer to computational techniques that can find a way to connect a set of inputs to a desired set of outputs by learning relevant data.
Artificial Intelligence (AI): when machines or software can perform tasks which usually require human intelligence, e.g. problem solving.
Bias: In AI models refers to output errors caused by skewed training data. Such bias can cause models to produce inaccurate, offensive, or misleading predictions. Biased AI models arise when algorithms prioritize irrelevant or misleading data traits over meaningful patterns (Smith, 2019).
Chatbot: A software program that is designed to simulate human conversation, often through text interactions.
Chain-of-thought Prompting: Chain-of-thought prompting is when you use a series of intermediate reasoning steps to improve the accuracy and applicability of answers generated by LLMs (Bubeck et al., 2022).
Completion: The generated output of a generative AI in response to a given prompt.
Context window: The context window is the maximum number of tokens (words or parts of words) that an AI model can process and consider simultaneously when generating a response. It is essentially the “memory” capacity of the model during an interaction or task. Models with larger context windows can handle larger attachments/prompts/inputs and sustain “memory” of a conversation for longer (Fogarty, 2023).
Deep Learning: A type of machine learning that uses neural networks with usually three or more (i.e. deep) layers to understand complex patterns in datasets.
Generative AI (GenAI): Generative AI refers to deep-learning models that can take raw data and “learn” to generate statistically probable outputs when prompted.
Generative Pre-trained Transformer (GPT): A family of neural network models that uses the transformer architecture and is trained on large datasets to generate human-like output.
Hallucination: A phenomenon wherein a large language model perceives patterns or objects that are nonexistent or imperceptible to human observers, creating outputs that are nonsensical or altogether inaccurate.
Input data: Input data is data added to an artificial intelligence (AI) to explain a problem, situation, or request. Input data may be cleaned, labeled, and organized, or it may be raw data. Often an AI is asked to create or synthesize information based on input data and the AI’s algorithm (which itself might have been fine-tuned or trained using a separate data set) in order to create new data (“output data”).
Large Language Model: A deep learning model that has been trained on large amounts of data and understands and generates text in a human-like fashion.
Natural Language Processing (NLP): A field of AI that focuses on how computers can understand, interpret, and generate human language.
Neural Network: A method in artificial intelligence that teaches computers to process data by mimicking how the human brain processes information.
Output: The generated answer of a generative AI's response to a given prompt, also known as 'completion'.
Prompt: The user input to a generative AI system, which guides the AI to generate the desired output. It is often in the form of questions in natural language.
Prompt Engineering: The technique of designing and crafting the prompt to get the best possible output.
Tool (or interface): The software that a user uses.
Training: The process of a model automatically learning patterns based on data.
Transformer: A neural network that learns context and meaning by tracking relationships in sequential data, such as words in sentences.
References:
1. https://mitsloanedtech.mit.edu/ai/basics/glossary/
2. https://libguides.mssm.edu/ai/keyterminology
