Recognises a derived column less sensitive than its source
A derived column at a lower level than its source, no derivation rule.
Can AI do this?
AI does this well Most of the tested AI models master this. 3 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 93% | Mastered |
| gemini-3.5-flash-lite | Low cost | 8% | Not shown |
| gpt-6-luna | Low cost | 95% | Mastered |
| claude-haiku-5-5 | Low cost | 94% | Mastered |
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: The column at its source's level, or the rule that allows less
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: CO-007 customer_profile.birth_date derived from party.personal_code, internal Acceptable: restricted as its source, or a rule (year of birth only)
Also known as: Derived column less sensitive than its source with no rule
Related knobits
Builds on
All 16 knobits of Classify data.
Source: Review checklist 02-classification, row 2 (DMBOK; 27002)