Glossary category archive for

AI Model Concepts

Fine-Tuning

Fine-tuning is the process of taking a pre-trained foundation model and continuing to train it on a smaller, more specific dataset to adapt its behavior for a particular domain, task, or use case. Rather than training a model from scratch — which requires enormous resources — fine-tuning starts from a model that already understands language…
Read more >: Fine-Tuning
Weights

Weights are the the billions of numerical parameters that define how the model thinks and responds: the internal values a model develops during training that determine how it responds to any given input. They are the accumulated result of the model being exposed to enormous amounts of data and adjusting, over millions of iterations, to…
Read more >: Weights
Multimodal

Multimodal describes an AI system or model that can process and generate more than one type of input or output, such as text, images, audio, video, or code. A multimodal model is not limited to a single data format. It can reason across multiple types at once, connecting information from an image and a text…
Read more >: Multimodal
Grounding

Grounding is the practice of connecting an AI model’s outputs to a specific, verifiable source of information, such as a document, knowledge base, or dataset, so that responses are anchored to that source rather than generated purely from the model’s training. A grounded response can be traced back to something real. An ungrounded response can’t.
Read more >: Grounding
Inference

Inference is the process by which a trained AI model generates a response or prediction from a given input. When a user submits a prompt, uploads a document, or triggers an AI feature, the model runs inference: applying what it learned during training to produce an output. Training is how a model learns; inference is…
Read more >: Inference
Hallucination

Hallucination is when an AI language model generates output that is factually incorrect, fabricated, or unsupported by available evidence, while presenting it with the same confidence as accurate information. The model is not lying or guessing in the way a person might, but it’s doing what it was designed to do, predicting the most statistically…
Read more >: Hallucination
Deterministic

Deterministic describes a system or process that produces the same output every time it receives the same input, without variation. Given identical conditions, a deterministic system follows a fixed, predictable path to a fixed, predictable result. There is no probability involved because the outcome is fully determined by the inputs and the rules governing the…
Read more >: Deterministic
Probabilistic

Probabilistic describes a system, output, or process that produces results based on likelihood rather than certainty. A probabilistic system does not follow a fixed, predetermined path to a single correct answer. Instead, it operates on patterns, weights, and statistical relationships to arrive at the most likely outcome given the available inputs, which means the same…
Read more >: Probabilistic
Token

A token is the basic unit of text that an AI language model reads and generates. A token isn’t the same as a word. It can be a whole word, part of a word, a punctuation mark, or a space, depending on how the model breaks text down. As a rough guide, one token is…
Read more >: Token
NLP

Natural language processing is the field of computer science focused on enabling machines to understand, interpret, and generate human language. LLMs are currently the most advanced tools for doing NLP work.
Read more >: NLP
LLM

A large language model is a type of AI model trained on massive amounts of text data to understand and generate human language.
Read more >: LLM