Dustin Tran

re:characteristic functions. It's a neat property for calculating characteristic functions. But not sure if it'll help us calculate `log_prob`? Maybe it's useful for recognizing that the sum of random variables...

You're correct. Programmatically, it's implemented with operators such as `__gt__`.

Thanks for working on this! Yes, getting it to work on Travis is very important. It looks like it's still failing? I restarted the build.

It's true that it's not _technically_ the prior because it's not standard normally distributed. Rather, it's the variational distribution with parameters initialized randomly. Clarification of this in the notebook is...

Looking purely at this line: ``` mus = tf.stack( [neural_network(x, qW_0.sample(), qW_1.sample(), qb_0.sample(), qb_1.sample()) for _ in range(10)]) ``` This draws functions from the predictive distribution under samples from the...

Likes a NaN when using HMC applied to the inverse-softplus transform of Gamma. The inverse-softplus unconstrains the Gamma distribution, which enables proper exploration. However, at close enough values to 0,...

No need to put effort into. @axch has been working on this. Maybe he can reply on when something public-facing may appear in TF?

Gave you permission. Looks like you need to accept before assignments are available. FYI, if I recall correctly, @axch has mostly been working on the adaptive path length. For reference,...

`rbf` is written as Tensor-in Tensor-out, so scipy isn't recommended unless you're willing to fetch the input out of the graph as you write here. It would be great to...

That's an interesting suggestion. I can see that as a potentially useful utility. Can you raise this in TensorFlow? It seems like a feature request that should be considered upstream...