Add to release-notes.

Author: twiecki

RELEASE-NOTES.md

@@ -16,6 +16,7 @@
 * Added a new notebook demonstrating how to incorporate sampling from a conjugate Dirichlet-multinomial posterior density in conjunction with other step methods (see [#4199](https://github.com/pymc-devs/pymc3/pull/4199)).
 
 ### New features
+- New experimental JAX samplers in `pymc3.sample_jax` (see [notebook](https://docs.pymc.io/notebooks/GLM-hierarchical-jax.html) and [#4247](https://github.com/pymc-devs/pymc3/pull/4247)). Requires JAX and either TFP or numpyro.
 - `sample_posterior_predictive_w` can now feed on `xarray.Dataset` - e.g. from `InferenceData.posterior`. (see [#4042](https://github.com/pymc-devs/pymc3/pull/4042))
 - Added `pymc3.gp.cov.Circular` kernel for Gaussian Processes on circular domains, e.g. the unit circle (see [#4082](https://github.com/pymc-devs/pymc3/pull/4082)).
 - Add MLDA, a new stepper for multilevel sampling. MLDA can be used when a hierarchy of approximate posteriors of varying accuracy is available, offering improved sampling efficiency especially in high-dimensional problems and/or where gradients are not available (see [#3926](https://github.com/pymc-devs/pymc3/pull/3926))