Guy Moss

Refactor neural_nets model structure

1

After the new DensityEstimator interface is introduced, we have a confusing structure: - [ ] `neural_nets.density_estimators` should contain only one file for flows, containing all classes of flow-based density estimators...

Adapt SNRE to new DensityEstimator abstraction

4

It would be good to create a `RatioEstimator` to use for SNRE methods, similar to `DensityEstimator` abstraction already implemented. The API and broadcasting should be similar, so possibly this should...

Make code formatting easier and less painful

5

There seem to be different local versions of ruff for different contributors, and successive PRs from different contributors change the same files back and forth depending on local versions. Need...

Hi, thanks for maintaining this package! While working on a separate [PR](https://github.com/sbi-dev/sbi/pull/1269) in the `sbi` package, we noticed that MADEs cannot correctly estimate the first dimension of the input when...

Avoid code repetition for embedding nets tests

1

We have a bunch of PRs right now implementing new embedding nets (#1503, #1499, #1494, #1472) - each one adds a test of the embedding net api equivalent to the...

Refactor pairplot function

11

100% of survey responders stated they use the pairplot function, however right now the API is messy and opaque. It would be great to keep the plot aesthetic the same,...

## 🚀 Feature Request ### **Is your request related to a problem?** Currently, we only have a few diffusion samplers for trained score estimators (see #1438). In addition to implementing...

We should implement checks for when the input condition is out of distribution (i.e. the simulator model is misspecified relative to the ground truth data). Specifically, we can start by...

The current test suite takes a long time to run, and contains a lot of redundancies. This slows us down a lot in development and integration when ensuring even small...

Set maximum time for rejection sampling

7

## 🚀 Feature Request ### **Is your request related to a problem?** Almost all posterior sample with [`accept_reject_sample`](https://github.com/sbi-dev/sbi/blob/06f13a8eb0c83dd1abb1cf4c33685ead60d133ec/sbi/samplers/rejection/rejection.py#L192). This raises a warning that sampling will be slow if the acceptance...