The authoritative terminology directory for mastering Generative Engine Optimization (GEO), the future of search (AI Overviews), and Large Language Models (LLMs).
A metric that measures how frequently an AI model cites a specific brand or domain as a source for its information.
The limited resources AI bots allocate to crawl and index a website for vector databases.
When an AI model generates factually incorrect or nonsensical information with high confidence.
Google's generative AI feature that provides a synthesized answer at the top of search results.
The process of optimizing content to be the direct answer provided by AI-powered 'answer engines' like Perplexity or ChatGPT.
A part of neural network architecture that allows the model to focus on specific parts of the input data.
The conceptual representation of a brand within a search engine or AI's knowledge graph.
The maximum amount of information (tokens) an LLM can 'keep in mind' during a single conversation or retrieval step.
AI systems designed to simulate human conversation, serving as the interface for most generative search engines.
A mathematical formula used to determine how closely related two pieces of content are in a vector space.
Experience, Expertise, Authoritativeness, and Trustworthiness—Google's quality guidelines for evaluating content.
A uniquely identifiable object or concept (person, place, thing, or brand) that search engines and AI models can distinguish from others.
The process of linking different mentions of the same real-world entity into a single unified identity for the AI.
A secondary AI model designed to verify the claims made by a primary generator model.
The process of training a pre-existing model on a specific dataset to improve its performance in a niche area.
A new SEO sub-discipline focused on increasing visibility within generative AI research snapshots and AI Overviews.
A search engine that synthesizes information to provide direct answers instead of just a list of links.
Ensuring an AI model's output is based on verifiable facts from a specific, reliable source.
The amount of new, non-redundant information a page provides compared to other search results on the same topic.
The process of categorizing what a searcher actually wants to achieve (e.g., informational, transactional, navigational).
The ratio of factual, useful information to 'fluff' or filler text on a given page.
A database of entities and their relationships used by AI to provide context and verify facts.
An AI model trained on vast amounts of text data to understand, generate, and process human language.
A technique used by AI to detect the underlying emotional tone or bias of a piece of content.
A multidimensional mathematical space where an LLM stores the relationships between concepts as vectors.
Botfusions' proprietary method of reinforcing brand entities within the mathematical weights of a model's latent representation.
AI models capable of processing and generating information across multiple formats, such as text, image, and video.
A branch of AI that helps computers understand, interpret and manipulate human language.
An AI technique used to connect users' queries with concepts and websites, even if the exact words don't match.
A measurement of how well a probability model predicts a sample; in AI search, a low score means high confidence.
The practice of designing and refining inputs to an AI model to get more accurate or useful responses.
An AI architecture that uses logical chains of thought to solve complex problems or queries.
A technique used to train AI models by incorporating human feedback to reward desired behaviors.
A weight assigned by an AI system to determine how relevant a piece of content is for a specific query.
An architectural pattern that allows LLMs to retrieve relevant information from an external knowledge base before generating a response.
Structured data added to a website to help search engines and AI models understand its content and relationships.
Google's experimental search interface that integrates generative AI results directly into queries.
The foundation of related topics and entities that define a brand's expertise to an AI.
A term that is conceptually related to another, forming a cloud of relevant entities around a topic.
Search techniques that focus on the user's intent and the contextual meaning of terms rather than just keyword matching.
The explicit credit given by an AI to the specific webpage it used to generate its answer.
Content generated entirely or partially by AI models.
The basic unit of text processing for LLMs, typically representing a few characters or a single word.
The selection of the 'K' most relevant documents from a vector database to be used as context for an AI response.
The massive collection of text data used to train a Large Language Model.
The underlying deep learning architecture that powers modern LLMs like GPT and Gemini.
Elements like reviews, security badges, and verifiable contact info that increase an AI's confidence in a brand.
When a brand is named online without a hyperlink; highly valued by LLMs for gauging true brand entity popularity.
A specialized database that stores information as high-dimensional vectors for semantic similarity searches.
The process of converting words or sentences into numbers so that AI can calculate the similarity between different pieces of information.
A search methodology that uses mathematical distances between vector embeddings to find relevant content.
A search query that is resolved on the results page (or inside an AI chat) without the user clicking on any external website.
Algorithms are changing. Learn how your brand is perceived by AI and how it can be optimized for LLMs.
Contact Us