# HG changeset patch
# User Yoshua Bengio <bengioy@iro.umontreal.ca>
# Date 1275415324 14400
# Node ID 6f042a71be239e9b4dac9e2b1e2816fb5e2b967a
# Parent  b8e33d3d7f650e6e251ef0ceb6344db5504f91a2
todo done

diff -r b8e33d3d7f65 -r 6f042a71be23 writeup/nips2010_submission.tex
--- a/writeup/nips2010_submission.tex	Tue Jun 01 13:57:16 2010 -0400
+++ b/writeup/nips2010_submission.tex	Tue Jun 01 14:02:04 2010 -0400
@@ -90,8 +90,10 @@
 of {\em out-of-distribution} examples and of the multi-task setting
 (but see~\citep{CollobertR2008}). In particular the {\em relative
 advantage} of deep learning for this settings has not been evaluated.
-
-% TODO: Explain why we care about this question.
+The hypothesis explored here is that a deep hierarchy of features
+may be better able to provide sharing of statistical strength
+between different regions in input space or different tasks,
+as discussed in the conclusion.
 
 In this paper we ask the following questions: