Module 2.4.3 · Phase 2 Decision · Step 2.4 Metrics & ThresholdsCore
LLM Evidence Record
What did the LLM assessment establish, and at which layer?
For a large language model, such as a chatbot, this page gathers what the LLM steps of the Fairness Navigator found: how the model answers test prompts for each group, what its provider discloses, and what the deployed system decided for real people.
What it does
It is read-only. Where you had no access, it shows that as a limit on the evidence, never as a finding.
Why it matters
This step translates fairness definitions into a binding measurement design, specifying metrics, thresholds, disaggregation requirements, and statistical validation criteria. It ensures that fairness evaluation is consistent, reproducible, and methodologically defensible.
Where it sits
- Fairness Definition
- Populations & Intersections
- Bias Identification & Prioritization
- Metrics & Thresholds
- Fairness Measurement Design
- Robustness Testing
- LLM Evidence Record
How to use it
In the Fairness Navigator Beta
Work through the steps in order. Several modules open from the Navigator's steps and hand their results back.
On its own Live
Open 2.4.3 from the platform's Modules page and run it when your project needs it. Work is saved with the assessment.
After a Pulse reading Live
Start with a fast first read, then go deeper with the modules where it points.