1
Problem
2
Logged data
CSV, one row per round
or drop it here
3
Context distribution
or edit the template below.
4
Policy to evaluate
5
Monte Carlo rollouts M
How M is chosen
- Roll the policy out m₀ times in the simulator fitted from your logs.
- Resample those pilot rollouts B times with replacement, and compute the interval width from each resample.
- Measure how much the width varies: ρ̂ = m₀ · Var(width) / mean(width)².
- Set M = max(m₀, ⌈z²1−τ/2 · ρ̂ / ε²⌉). With M rollouts, the Monte Carlo error changes the interval width by less than a fraction ε, with probability 1 − τ.
- If that M is above your cap, BSI uses the cap and warns you that the Monte Carlo error may be larger than ε.
Advanced settings
6