Space Regulatory Glossary AI & Data Science

large language model

noun (abbr.)

1. A neural network trained on large text corpora that generates, summarizes, and reasons about natural language. Models like GPT-4, Claude, and Llama vary in size, cost, and suitability for different tasks.
2. In space regulatory workflows, LLMs are used for structured extraction from filings, classification of document types, and natural-language querying of regulatory databases. Performance varies significantly by domain, making benchmarking against ground-truth data essential before production use.

Read: Benchmarking LLMs for Domain-Specific Extraction

Related terms

retrieval-augmented generation An AI architecture that grounds a language model's output in retrieved source documents rather than relying solely on… fine-tuning The process of further training a pre-trained AI model on a domain-specific dataset to improve its performance on… inference The process of running a trained AI model to produce outputs from new inputs, as opposed to training the model.… hallucination An AI model output that is factually incorrect, fabricated, or unsupported by the source data, presented with the same…

Mentioned in The Downlink

• AI Verification Engineering: Architecture for Trustworthy Regulatory AI

  RAG hallucinates at 17 to 33 percent in production legal tools. The architecture of trustworthy regulatory AI: source grounding, extraction validation, confidence scoring, and standards-aligned verification.

Browse all terms in the Space Regulatory Glossary →