Explains metric direction
Which way is good must follow from the definition; a count that rises with problems and with honest reporting has no readable direction.
Can AI do this?
AI does this partly Some of the tested AI models master this; others do not show it. 2 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 99% | Mastered |
| gemini-3.5-flash-lite | Low cost | 29% | Not shown |
| gpt-6-luna | Low cost | 97% | Mastered |
| claude-haiku-5-5 | Low cost | 65% | Not shown |
Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.
What it looks like in practice
Expected result: Which way is good must follow from the definition; a count that rises with problems and with honest reporting has no readable direction.
How it is assessed: Answers a written question on the metric direction correctly
Related knobits
Needed for
- G-608 Recognises a metric whose direction cannot be read
- G-632 Replaces an unreadable metric direction with readable metrics
All 21 knobits of Build the governance scorecard.
Source: Practice the stage 6 reference and review checklist rest on