Dingkun.Liu

icon company Tsinghua University icon location Beijing, China icon location PhD student of Lab of Neural Engineering, Tsinghua University.

Results 3 comments of Dingkun.Liu

q3_gru.py add_prediction_op() 中dynamic_rnn应该取output作为值而不是state吧?

Vanilla RNN中也是这样的，这里应该是不考虑LSTM的。

Questions about robustness

In most cases, XdawnCovariances works better for me.

为啥生成的书签，后面都会加一个 -0.2em?

> delete "\vspace{-0.2em}" in line 79: > \def\cvprsubsect#1{\cvprsubsection{\hskip -1em.~#1 \vspace{-0.2em}}} 你的花括号位置错了，应该是@KimHe的方案 > change it to > `\def\cvprsubsect#1{\cvprsubsection{\hskip -1em.~#1 }\vspace{-0.2em}}`