Recognises a correction loaded as an unrelated fact
A reversal loaded with no link to its original.
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 | 83% | Not shown |
| gemini-3.5-flash-lite | Low cost | 14% | Not shown |
| gpt-6-luna | Low cost | 83% | Not shown |
| claude-haiku-5-5 | Low cost | 83% | 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 flow and the link to the original
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: IF-002 reversal events inserted as new transactions Acceptable: reverses: original_txn_id, netted in reporting
Also known as: Correction as an unrelated new fact
Related knobits
Builds on
Usually followed by
All 15 knobits of Design the integration patterns.
Source: Review checklist 03-integration-patterns, row 4 (DMBOK)