I was trying to following this example implemented in Pyro,
while setting up the prior distribution,
prior = dist.Beta(10, 10)
plt.figure(num=None, figsize=(10, 6), dpi=80)
x_range = np.linspace(0, 1, num=100)
y_values = torch.exp(prior.log_prob(torch.tensor(x_range)))
I am not clear why we have to compute log_probability first and then apply exponential to generate y_value.