A score tells you what. Not why.
The same wrong answer can come from different mental models. Each needs a different explanation; repeating the answer key may fix neither.
Most quizzes tell you which answer was right. Faultline reads why you chose the one you did, finds the exact misconception behind it, and teaches only that.
Same red cross, opposite fixes. A score cannot see the difference.
the question that splits the remaining faults
your answer, how sure you were, and your own words
the fault underneath, and nothing else
The same wrong answer can come from different mental models. Each needs a different explanation; repeating the answer key may fix neither.
Use it as a formative learning aid to explore mechanics misconceptions and choose a next teaching step. It is not a grade, diagnosis, or replacement for a teacher.
Today’s models can explain almost anything; a fluent answer is not a diagnosis. A Bayesian engine chooses questions and weighs evidence. Optional Neural Pulse can read free-text reasoning and reframe authored teaching, but cannot choose a fault or invent the lesson.
No key: reasoning and profiles stay in this browser. Connected mode sends written reasoning plus question context to Neural Pulse and syncs profile/session data, including this label. Avoid entering names or sensitive details in free-text responses.