Main Page

llmref.wiki is a neutral, technical reference for LLM-era (post-2022) terminology — generative search, AI visibility, prompt engineering, agents, and evaluation.

Definitions here are source-first and, wherever a term is commonly confused with another, explicitly distinguished from it. The editorial standard is Wikipedia-style neutrality with W3C-style precision: definitions and methodology only, no marketing language, primary sources cited.

Browse by category

• Fundamentals — core LLM concepts (context window, tokens, hallucination)
• Generative search — GEO, AEO, LLMO, AI SEO
• AI visibility — AI visibility, Citation rate vs Mention rate, Share of voice in AI, Entity authority
• Measurement — Source attribution, citation accuracy
• Prompt engineering — Grounding vs RAG, Prompt injection vs Jailbreak
• Agents — Agentic AI vs AI agent, Agent memory vs Context window, Tool use vs Function calling
• Evaluation — Faithfulness vs Groundedness, Benchmark contamination
• Standards — llms.txt, AI crawler, E-E-A-T in the AI era

Terminological conflicts this reference resolves

Many LLM-era terms have no authoritative source distinguishing them from near-synonyms. Pages that disambiguate competing definitions include:

• GEO vs AEO vs LLMO vs AI SEO
• Generative engine vs Answer engine
• Citation rate vs Mention rate
• Grounding vs RAG
• Prompt injection vs Jailbreak
• Agentic AI vs AI agent
• Agent memory vs Context window
• Tool use vs Function calling
• Faithfulness vs Groundedness

About

This is a reference project. Content is maintained to a fixed editorial standard; see individual pages for sources. For machine consumption, see llms.txt.