Recognises a computed result taken on trust
A computed result credited without recomputing it from the task data.
Can AI do this?
Not settled yet The benchmark tested this, but the models' answers do not settle yet whether they master it. 0 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 79% | Not shown |
| gemini-3.5-flash-lite | Low cost | 79% | Not shown |
| gpt-6-luna | Low cost | 42% | Not shown |
| claude-haiku-5-5 | Low cost | 79% | 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: The recomputed value and the score it earns
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: B's "(310 + 90 + 860) / 3,000 = 38%" credited Acceptable: Recomputed: 42.0%; C3 = 0
Also known as: Arithmetic taken on trust
Related knobits
Builds on
All 16 knobits of Grade answers against a rubric.
Source: Review checklist 03-answer-grading, row 10 (Standards 6)