Three questions/problems after extensive using #1981
Replies: 5 comments
-
The second problemI think you should use double As your example, $$
\begin{alignat}{2}
v_{\pi 1}^{(0)} &= v_0 \nonumber\\
v_{\pi 1}^{(1)} &= r_{\pi 1} + \gamma P_{\pi 1} v_{\pi 1}^{(0)} &&\longrightarrow v_1 \longrightarrow \text{Value Iteration} \nonumber\\
v_{\pi 1}^{(2)} &= r_{\pi 1} + \gamma P_{\pi 1} v_{\pi 1}^{(1)} \nonumber\\
\vdots \nonumber\\
v_{\pi 1}^{(j)} &= r_{\pi 1} + \gamma P_{\pi 1} v_{\pi 1}^{(j-1)} &&\longrightarrow \bar{v_1} \longrightarrow \text{Truncated Policy Iteration} \nonumber\\
\vdots \nonumber\\
v_{\pi 1}^{(\infty)} &= r_{\pi 1} + \gamma P_{\pi 1} v_{\pi 1}^{(\infty)} &&\longrightarrow v_{\pi 1} \longrightarrow \text{Policy Iteration} \nonumber
\end{alignat}
$$ |
Beta Was this translation helpful? Give feedback.
-
@huanyushi |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
The third problemIt seems that $$
\begin{align}
D_{KL}(P^*(y|x_i)\ ||\ P(y|x_i; \Theta))
&= \sum_{y} P^*(y|x_i) log \frac{P^*(y|x_i)}{P(y|x_i; \Theta)} \\
&= \sum_{y} P^*(y|x_i)[logP^*(y|x_i) - logP(y|x_i; \Theta)] \\
&= \sum_{y} P^*(y|x_i)logP^*(y|x_i) - \sum_{y} P^*(y|x_i)logP(y|x_i; \Theta)
\end{align}
$$ |
Beta Was this translation helpful? Give feedback.
-
@huanyushi For the 3rd problem, changing code block type from markdown to latex can solve the problem with |
Beta Was this translation helpful? Give feedback.
-
Checklist
How did you create the site?
Generated from
chirpy-starter
Describe the bug
I have made a post to reproduce all problems described in this issue here.
Verticsl scroll bar appeared in list when it shouldn't
Math euqation alignment at multiple palces
Here is the markdown for the equation
Code block grammar check (switching themes between dark and light is related)
Example 1
In light mode
Here is the HTML element for the code block
In dark mode,
Here is the HTML element for the code block
Example 2
The second problem is simpler. The standard output from a python console was marked problematic here.
Steps To Reproduce
I have made a post to reproduce all problems described in this issue here and you can find the markdown file for the post here.
Expected Behavior
Environment
Anything else?
No response
Beta Was this translation helpful? Give feedback.
All reactions