Using GenJax 0.1.1 in JupyterLab 4.3.4 in Python 3.13 on MacOS 15.5, the example of graphing the means of the estimated values surprisingly does not converge to the true value even when 100 million samples are generated, but instead seems to converge to about 0.117 versus the true value 0.105.
Perhaps as with the failure of AD operation later on in the example, this is an artifact of randomness in the model, but since the page says, "As we use more and more samples we will converge to the correct result by the Central limit theorem," it seems that this isn't expected, so I just thought I'd mention it in case there could be a bug. The issue seems to persist regardless of the value used to initialize the random key. Odd.