# Explanation

Understanding-oriented essays, one concept per page: what simulation-based
inference is and the simulate → train → validate → export workflow, why a
learned likelihood is validated before it is trusted, why hierarchical
inference is powerful, and how fast setu is on a single measured task.

```{toctree}
:maxdepth: 1

What is SBI? The four-phase workflow <what-is-sbi>
Why validate a learned likelihood <why-validate>
Why hierarchical inference is powerful <why-hierarchical>
How fast is setu, and on what <benchmarks>
```