GEO
LLM answer
An LLM answer is the text response returned by a declared large-language-model endpoint or product interaction for a specific input and available context.
What it means
The label identifies a model-mediated response only when the model or product documentation supports that Attribution. A product may add search, retrieval, policy filters, tools or post-processing, so calling the whole experience a raw model answer can be inaccurate. Model name, version, system context and tool availability may be known, partially known or unavailable and must retain those states.
Why it matters
Separating model output from the surrounding product prevents teams from attributing sources, rankings or interface behavior to an LLM alone.
Example
An evaluation stores a response from a documented model endpoint with model version, input, system instructions, temperature and tool setting; an assistant product capture is labelled product output when those fields are hidden.
Common mistakes
Do not guess the model from prose, call every assistant response raw LLM output, ignore retrieval or tools, treat one response as deterministic, infer training sources or publish hidden prompts.
How AYSA handles this
Signals reviewed
provider documentation, model identifier, endpoint or product, input, system context, tool state, sampling, output
Problem AYSA can identify
AYSA can identify guessed model attribution, product outputs mislabelled as raw model responses and unavailable settings converted into assumed values.
Recommendation prepared
The workflow selects the narrowest defensible attribution and preserves known, unknown and not-applicable fields.
Approval preview
The reviewer sees provider evidence, model or product scope, input disclosure, tool state, unavailable fields and prohibited training-source claims.
Execution
AYSA can evaluate authorized model or product responses; it cannot recover hidden prompts, routing, weights or training data.
Verification
Each rerun retains its model or product context and output rather than being averaged into a fictional canonical answer.
Limits
Model attribution may remain incomplete, and logs or human review do not automatically certify AI Act or legal compliance.
Sources and further reading
- NIST — Artificial Intelligence Risk Management Framework: Generative AI Profile — Official risk-management framework
- W3C — PROV Overview — W3C provenance standard
- Enabling Large Language Models to Generate Text with Citations — Original academic research
- European Commission — AI Act Article 50 transparency guidelines — Official European Commission guidance
Written by Marius Dosinescu. Reviewed by AYSA SEO Editorial Team · 2026-07-28 00:00:00.
Quick answers
Frequently asked questions
Is every assistant response a raw LLM answer?
No. Products may add retrieval, tools, filters and post-processing around a model.
Can an LLM answer reveal its training sources?
No. Visible wording or citations do not disclose the model's training corpus.
SEO execution campaign
Less SEO work. More organic growth.
AYSA monitors your website, finds opportunities, prepares the work, asks for approval and executes accepted changes so you can grow without living in SEO tools.
