Updates on each answer with the item's own parameters
From a uniform prior over the states, each answer multiplies every state's probability by the item's likelihood (1 - beta or beta when the item is solvable in the state, eta or 1 - eta when not); a guessable item's right answer counts for less.
Can AI do this?
Not tested yet
No AI benchmark covers this knobit yet. New results are announced in the Can AI Do the Job? newsletter.
All 7 knobits of Diagnose a learner on a knobit graph.