The AI & ML
Dictionary

Accuracy measures how often an AI system's predictions or outputs are correct compared to the actual truth. It's calculated as the percentage of corre...

Agentic Workflow refers to AI systems that don't just answer questions but actively perform tasks. Instead of a single prompt-response loop, an agenti...

An AI agent is a computer program that can perceive its environment and take actions to achieve specific goals. Think of it as a digital assistant tha...

An AI API is a programming interface that allows developers to access artificial intelligence capabilities without building the AI models themselves....

A benchmark is a standard test used to measure and compare the performance of different AI systems. It helps developers understand which AI model work...

AI bias occurs when artificial intelligence systems produce unfair or prejudiced results that favor certain groups over others. This happens because t...

An AI chatbot is a computer program that can have conversations with people using natural language. It understands what you type or say and responds i...

An AI Copilot is an artificial intelligence assistant that helps people complete tasks by providing real-time suggestions and guidance. It works along...

Code Completion AI is artificial intelligence that helps programmers write code faster by automatically suggesting the next lines or completing code s...

Context Caching is a technique that lets developers 'save' the state of a massive prompt (like a whole book or codebase) so they don't have to pay to...

A context window is the amount of text or information an AI model can consider at one time when generating responses. It's like the model's short-term...

Conversational AI refers to artificial intelligence systems designed to have natural, human-like conversations with people. These systems can understa...

A dataset is a collection of information used to train AI systems. Think of it like a textbook that teaches AI how to recognize patterns and make deci...

Deep Learning is a type of artificial intelligence that teaches computers to learn from examples, much like how humans learn from experience. It uses...

A diffusion model is an AI technique that creates images by starting with random noise and gradually refining it into a clear picture. It works by lea...

Embeddings are a way to convert words, sentences, or other data into lists of numbers that computers can understand. These number lists capture the me...

Explainability refers to how well we can understand why an AI system makes a particular decision or produces a specific output. It's about being able...

Few-shot learning is an AI technique where a model learns to perform new tasks with very few examples. Instead of needing thousands of training sample...

Fine-tuning is the process of taking a pre-trained AI model and adapting it to perform better on a specific task or domain. It's like taking a general...

A GAN is an AI system where two neural networks compete against each other - one creates fake content while the other tries to detect what's fake. Thi...

GPT is a type of artificial intelligence that can generate human-like text by learning from vast amounts of written content. It works by predicting th...

Guardrails are safety measures that prevent AI systems from generating harmful, inappropriate, or factually incorrect content. They work like digital...

Hallucination occurs when an AI system generates information that seems plausible but is actually incorrect or made up. This happens because AI models...

Image Generation AI refers to artificial intelligence systems that can create new images from text descriptions or other inputs. These systems learn f...

Inference is when an AI model uses what it has learned to make predictions or generate responses based on new input. It's like the AI putting its trai...

Jailbreaking is when someone finds clever ways to make an AI system ignore its safety rules and restrictions. It's like convincing a very smart assist...

Latency is the delay between when you send a request to an AI system and when you receive a response. It measures how long you have to wait for the AI...

A Large Language Model is an AI system trained on massive amounts of text data to understand and generate human-like language. It can read, write, tra...

Machine Learning is a type of artificial intelligence that enables computers to learn and improve from experience without being explicitly programmed...

Model training is the process of teaching an AI system to perform specific tasks by showing it many examples. It's like teaching a child to recognize...

MoE is a model architecture that divides an AI into many smaller, specialized 'expert' sub-networks. Instead of activating the entire massive brain fo...

Multimodal AI refers to artificial intelligence systems that can understand and process multiple types of information simultaneously, such as text, im...

A neural network is a computer system designed to mimic how the human brain processes information. It learns patterns from data to make predictions or...

Overfitting happens when an AI model learns the training data too well, including its random noise and specific details, instead of the general patter...

Precision measures how accurate an AI model is when it makes positive predictions. It tells you what percentage of the model's 'yes' answers were actu...

Prompt engineering is the practice of carefully designing and refining the instructions given to AI systems to get better results. It involves craftin...

RAG is an AI technique that combines information retrieval with text generation. It first searches through documents or databases to find relevant inf...

A Reasoning Model is an AI designed to 'think' before it speaks. Unlike standard chatbots that predict the next word instantly, these models generate...

Reinforcement Learning is a type of AI training where an agent learns by interacting with its environment and receiving rewards or penalties for its a...

RLHF is a training method where AI models learn to produce better responses by getting feedback from humans. Instead of just learning from data, the m...

Speech-to-Text is a technology that converts spoken words into written text. It allows computers to understand and process human speech, making it pos...

Supervised learning is a type of AI training where the computer learns from labeled examples, like a student learning with answer keys. The AI studies...

Synthetic Data is information artificially generated by AI to train other AI models, rather than collected from the real world. As we run out of high-...

Temperature is a setting that controls how random or predictable an AI's responses will be. A low temperature makes the AI more focused and consistent...

Text-to-Speech is an AI technology that converts written text into spoken audio. It allows computers to read text aloud using synthetic voices, making...

A token is the basic unit of text that AI models process, representing words, parts of words, or even individual characters. Think of tokens as the bu...

Top-p sampling is a method used by AI systems to choose which words or tokens to generate next. Instead of always picking the most likely option, it c...

Transfer learning is when an AI model uses knowledge it learned from one task to help it perform a different but related task. Instead of starting fro...

A transformer model is a type of AI architecture that processes information by paying attention to different parts of the input simultaneously. It rev...

Underfitting occurs when an AI model is too simple to capture the patterns in the data it's learning from. This means the model performs poorly on bot...

Unsupervised learning is a type of AI that finds patterns in data without being told what to look for. It works by grouping similar items together or...

An AI video generator is a tool that creates videos automatically using artificial intelligence. It can turn text descriptions, images, or other input...

AI Voice Cloning is a technology that creates a digital copy of a person's voice using artificial intelligence. It can generate new speech that sounds...

A vector database is a specialized type of database that stores data as mathematical vectors instead of traditional text or numbers. It's designed to...

An AI writing assistant is a computer program that helps people create and improve written content. It can generate text, suggest edits, check grammar...

Zero-shot learning is an AI approach where a model can recognize or understand things it was never specifically trained on. It works by learning gener...