Misconception Detectionpass

A model-assisted pass that proposes common misconceptions about a concept and their corrections, attached as MisconceptionNodes.

A concept is not understood until its common failures are named. Proactive correction compiled in beats a chatbot guessing at query time.

Where it appears

Descend

Historical evolution

Listed as a future compiler direction in the project brief; implemented as a local-model enrichment pass.

Implementation details

Every MisconceptionNode carries confidence + source; human-review queue before publish; never presented as established fact.

Sources

View in graph →