Generative AI and LLM glossary

Retrieval

RAG, retrieval-augmented generation, retrieves evidence first, then generates an answer from that evidence instead of parametric memory alone.

Inference

Reasoning effort is a model setting or design choice that controls how much reasoning budget the model spends before answering.

Disaster Recovery

A metric that defines the maximum acceptable amount of data loss measured in time for a system after a failure.

Deep Learning

A type of neural network designed for sequential data processing.

Modeling

A technique used to prevent overfitting by adding a penalty to the loss function.

Machine Learning

A type of machine learning focused on teaching agents to make decisions by maximizing cumulative rewards.

Training

RLHF, reinforcement learning from human feedback, trains a model to prefer outputs that raters rank higher, usually via a reward model.

Automation

Robotic process automation (RPA) is the use of software robots to automate repetitive tasks.

Engineering

The interdisciplinary field that combines engineering, computer science, and more to design and build robots.

Quality Assurance

Evaluation processes to ensure the functionality and safety of robotic systems.

Model Evaluation

The ability of a model to maintain performance despite variations in input data or conditions.

Model Evaluation

Evaluating how well a model performs under various adversarial conditions.

Data Analysis

The process of identifying the primary cause of a problem or defect.