
LearnBook's learning loop connects importing and reading to testing and review. Here is how LearnBook's review and retention layer works today. The focus is a single question: once you have studied something, how does LearnBook help you keep it? What follows is a feature-by-feature look at the review and retention system. This is part of our ongoing product updates series.
Reviews are now scheduled using FSRS (Free Spaced Repetition Scheduler), a modern algorithm that models how memory decays over time for each individual card. Rather than showing you every card on a fixed daily cycle, FSRS predicts the optimal moment to review each item — early enough that you still remember it, late enough that the review reinforces the memory most effectively.
In practice, this means your review sessions get shorter as you demonstrate mastery, and items you find difficult surface more frequently until they stabilize. The algorithm adapts to your actual performance, not a one-size-fits-all schedule. For a fuller explanation of the mechanics, see how FSRS works and what spaced repetition is.
Knowledge rarely lives in a single source. A concept introduced in one book connects to examples in another; a framework you learned in a lecture gets reinforced by a paper you read later. Cross-book review surfaces flashcards and quiz items from across your entire library in a single session, so the connections between your sources become part of how you review — not something you have to manage manually.
As your library grows, the review session draws from a wider pool, which means ideas you learned months ago remain active alongside material you just imported.
The mastery trends view gives you a per-topic picture of how your retention is evolving over time. Rather than a single pass/fail grade on a quiz, mastery trends track your recall rate for cards and questions grouped by concept and source, letting you see which areas are solidifying and which are slipping.
This is especially useful for identifying the topics that feel familiar but are not yet reliable under retrieval pressure — the gap between recognition and actual recall that spaced repetition is designed to close.
For newly imported material, the 21-day memory loop structures your reviews around the intervals that research identifies as most critical for moving information from short-term exposure into longer-term retention. The loop does not require any manual setup: once you have generated flashcards or quizzes from a source, the schedule begins automatically.
The goal is to reduce the effort cost of consistent review. Rather than relying on willpower to decide when to revisit material, the 21-day loop handles the scheduling so that the right items appear at the right times during the period when forgetting is most likely to occur.
Consistency matters more than intensity in long-term learning. Study streaks track your daily review activity and reward continuous engagement, giving you a visible measure of momentum that is worth protecting.
The XP and level system accumulates across all your review activity — flashcards, quizzes, and daily sessions — translating effort into a persistent record of progress. Levels reflect your overall engagement with the system, not just performance on any single topic.
Learning does not have to be solitary. Team leaderboards let groups — study cohorts, classrooms, or colleagues working through the same material — compare activity and review streaks, adding a layer of social accountability.
Quiz challenges let you compete on specific topics, turning review into something closer to a game. Friend PK (head-to-head) matches you against another user for a timed quiz session on shared material, which adds a different kind of pressure than solo review — useful for stress-testing knowledge you think you have mastered.
These social features are optional. The review system works entirely on its own if you prefer to study alone.
The review and retention layer described here sits on top of the full Learning Loop — Import, Compile, Read & Ask, Quiz / Flashcards, Review — without changing the earlier stages. Flashcards are auto-generated from your sources with citations. The AI Tutor grounds every answer in your own material. The loop ends with a review session; the features above determine how deep and how personal that final stage is.
If you have material you have already imported and studied, the review system is ready to pick up where you left off. If you are new to LearnBook, bring in something you genuinely want to retain long-term and see how the full loop holds up over weeks, not just a single session. For the original announcement of the full learning loop, read the introduction to LearnBook's learning loop.
Import, compile, read, quiz, review — the LearnBook learning loop turns any material into verifiable learning. Here's the full picture.