Recognises weights deciding the outcome alone
Weights not summing to 100, or set after scoring, or one theme counted so heavily that the scores no longer matter.
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 | 94% | Mastered |
| gemini-3.5-flash-lite | Low cost | 8% | Not shown |
| gpt-6-luna | Low cost | 94% | Mastered |
| claude-haiku-5-5 | Low cost | 95% | 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 defect named on the weights not set or not summing before scoring, and its fix or the question for its owner
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: Weights sum to 115; cost appears twice (40 % of the total) Acceptable: Weights sum to 100, agreed before scoring, one criterion per theme
Also known as: Weights that decide alone
Related knobits
Builds on
- O-151 Explains gated weighted evaluation
- Q-1804 Explains the concept of a percentage
- O-1902 Explains a weighted score as the sum of weights times scores
All 14 knobits of Evaluate data storage technology.
Source: Review checklist 01-technology-evaluation, row 5 (MCDA)