This tool uses paid AI to read your PDFs. Enter the password to continue.
Load a list of exercises, set your study and rest times, and how many cycles you want. A cycle is one study block followed by one rest block.
Upload a PDF and I'll extract the questions. Extraction is never perfect, so you'll review and edit them below before starting. You can also skip the PDF and type questions in by hand.
Fix the split if needed. Each question can have an answer — the trainer reveals it during the session, then you mark whether you got it right. If a question has no answer, you just self-assess.
Did you get it right?
Cycle 1 done. Next study block starts when the timer ends.
Dots are your discrete data points (green = right, red = wrong). The curve is the least-squares fit of S(t) = a(1 − e^−kt). This form is a working hypothesis: it follows if the answering rate decays exponentially with fatigue, r(t) = r₀·e^−kt, integrating to S(t) = (r₀/k)(1 − e^−kt), so a = r₀/k. It is coherent with the model's fatigue form, not proven — the data is what tests it.