reb

Author: cdluminate

reb/RebuttalTemplate.tex

@@ -219,16 +219,15 @@
 The PGD complexities per adversarial
 example are identical across different defenses, and hence omitted.
 %
-Triplet sampler has been aligned to ``Softhard'' for identical complexities, and
-hence omitted.
+Triplet sampler is aligned to ``Softhard'' in the table for identical
+sampling complexities, and hence omitted.
 %
-The rest complexities stem from the adversarial perturbation objective functions,
-as reported in table.
+The rest complexities stem from adversarial perturbation objective functions.
 %
-Note, the reported values are sensitive to hardware, software, and
+Note, the values are sensitive to factors including hardware, software, and
 programming tricks, which makes them uneasy to compare for future works.
 %
-These information will be provided in the next revision.
+The comparison will be provided.
 
 %\subsubsection*{To Reviewer \#3}
 \vspace{0.3em}
@@ -286,7 +285,7 @@
 example performance than previous state-of-the-arts~\cite{advrank,robrank}.
 
 %%%%%%%%% REFERENCES
-\vspace{-0.5em}
+%\vspace{-0.5em}
 {\small
 \bibliographystyle{ieee_fullname}
 \bibliography{egbib}