Simulated learner run
Test the adaptive policy before a learner sees it.
This page runs the probability/concentration gold bridge set against fixed synthetic learner profiles. It checks whether the policy routes learners through prerequisite claims before sending them to finite-class generalization.
Gold set
bridge v1
Profiles
4
Claims covered
8
Steps per run
10
Gold set
Probability / concentration bridge gold set v1
A curated 10-question diagnostic bridge from sub-Gaussian tails through finite-class uniform convergence.
probability-concentration-bridge-v1
| Synthetic profile | Prerequisite-aware policy | Jumpiest baseline |
|---|---|---|
Probability basics gap probability-basics-gap Best raw accuracy: Prerequisite-aware at 100%. | Prerequisite-aware 0 prerequisite jumps accuracy 100% mastery 43% Next checkpoint: probability-event-algebra-and-moments First pick: question:diverse-state-theorem-001 | Explore/exploit 7 prerequisite jumps accuracy 80% mastery 43% Next checkpoint: probability-event-algebra-and-moments First pick: question:logic-ml-foundations-030 |
Concentration assumptions gap concentration-assumptions-gap Best raw accuracy: Static path at 80%. | Prerequisite-aware 0 prerequisite jumps accuracy 40% mastery 42% Next checkpoint: probability-event-algebra-and-moments First pick: question:diverse-state-theorem-001 | Explore/exploit 6 prerequisite jumps accuracy 50% mastery 42% Next checkpoint: probability-event-algebra-and-moments First pick: question:diverse-state-theorem-001 |
Uniform convergence bridge gap uniform-convergence-bridge-gap Best raw accuracy: Prerequisite-aware at 70%. | Prerequisite-aware 0 prerequisite jumps accuracy 70% mastery 37% Next checkpoint: probability-event-algebra-and-moments First pick: question:mcq-fundamental-theorem-012 | Explore/exploit 7 prerequisite jumps accuracy 60% mastery 37% Next checkpoint: probability-event-algebra-and-moments First pick: question:logic-ml-foundations-030 |
Rusty probability theorist rusty-probability-theorist Best raw accuracy: Explore/exploit at 90%. | Prerequisite-aware 0 prerequisite jumps accuracy 60% mastery 55% Next checkpoint: probability-event-algebra-and-moments First pick: question:subgaussian-sum-tail-bound-012 | Review due first 5 prerequisite jumps accuracy 40% mastery 54% Next checkpoint: probability-event-algebra-and-moments First pick: question:subgaussian-sum-tail-bound-012 |
Review focus
- sub-Gaussian MGF assumptions
- Hoeffding versus Chebyshev scope
- finite union bound in learning theory
- uniform convergence prerequisites
Project unlocks
- confidence-radius-widget
- finite-erm-sample-size-planner
Interpretation
This simulation validates routing behavior on fixed profiles. It does not estimate real learner ability. Real calibration starts after answer events are collected.