user8714896
  • Member for 2 years, 1 month
  • Last seen more than a month ago
2 answers
5 votes
364 views
Why is "-0.5 * torch.sum(1 + sigma - mu.pow(2) - sigma.exp())" in Pytorch equivalent to the KL?
1 answers
1 votes
71 views
How can max-pooling be applied to find features in words?
2 answers
5 votes
7k views
4 bookmarks
What exactly is a hidden state in an LSTM and RNN?
1 answers
1 votes
422 views
How to alpha-beta pruning to expectiminimax
1 answers
0 votes
157 views
Can you train Transformers sequentially?
0 answers
2 votes
26 views
The MLP output of a neural network can be written as $\|x\|\|w_l\|\cos(\theta_l)$: why is the norm easier to maximize?
1 answers
2 votes
130 views
Is there a mathematical theory behind why MLP can classify handwritten digits?
0 answers
0 votes
44 views
How to implement base cases of C4.5 decision tree algorithm?
0 answers
1 votes
10 views
Can you correlate decision boundary of final layer of a neural network to predictive distribution?
0 answers
3 votes
80 views
Why does this formula $\sigma^2 + \frac{1}{T}\sum_{t=1}^Tf^{\hat{W_t}}(x)^Tf^{\hat{W_t}}(x_t)-E(y)^TE(y)$ approximate the variance?
1 answers
2 votes
75 views
How UCT in MCTS selection phase avoids starvation?
1 answers
0 votes
117 views
Does Monte Carlo Tree Search not work on games without the same initial state?
0 answers
0 votes
32 views
Beating iterative alpha beta search in Isolation Game
1 answers
1 votes
36 views
Embedding Isolation game states into key values for RL
1 answers
3 votes
366 views
Why does every neuron in hidden layers of a multi-layer perceptron typically have the same activation function?
1 answers
2 votes
63 views
1 bookmarks
Is there any research on models that provide uncertainty estimation?
1 answers
2 votes
74 views
Why is neural networks being a deterministic mapping not always considered a good thing?
1 answers
9 votes
291 views
3 bookmarks
How does the forget layer of an LSTM work?
1 answers
1 votes
69 views
If $\alpha$ decreases over time, why is Q-learning guaranteed to converge?
0 answers
0 votes
42 views
Suppose every-visit MC was used instead of first-visit MC on blackjack. Would you expect the results to be different?
1 answers
2 votes
47 views
Why is it useful to define the return as the sum of the rewards from time $t$ onward rather than up to $t$?
1 answers
1 votes
46 views
For episodic tasks with an absorbing state, why can't we both have $\gamma=1$ and $T= \infty$ in the definition of the return?
0 answers
3 votes
437 views
Why would a VAE train much better with batch sizes closer to 1 over batch size of 100+?
1 answers
1 votes
69 views
How does implementation of VAE's objective function equate to ELBO?
1 answers
1 votes
78 views
In this VAE formula, why do $p$ and $q$ have the same parameters?
1 answers
4 votes
309 views
1 bookmarks
Why is the evidence equal to the KL divergence plus the loss?
1 answers
2 votes
47 views
Why do we need to go back to policy evaluation after policy improvement if the policy is not stable?
1 answers
2 votes
303 views
1 bookmarks
Why does KL divergence not satisfy the triangle inequality?
0 answers
4 votes
149 views
1 bookmarks
Why can't VAE do sequence to sequence name generation?
0 answers
1 votes
18 views
Creating a noising model for NLP that models human noising