97 lines
1.0 KiB
Plaintext
97 lines
1.0 KiB
Plaintext
<subsection For few Particles>
|
|
|
|
|
|
<frame title="Does it actually Autoencode?">
|
|
First try this for tiny graphs of about 4 particles
|
|
<split>
|
|
<que>
|
|
<i f="history4" wmode=True>loss per epoch (4 nodes)</i>
|
|
</que>
|
|
<que>
|
|
<i f="reloss" wmode=True>comparison of the loss to trivial reconstructors</i>
|
|
</que>
|
|
</split>
|
|
|
|
</frame>
|
|
|
|
<ignore>
|
|
dont really have nice images for this
|
|
<frame>
|
|
|
|
<split>
|
|
<que>
|
|
<i f="none">a good reconstruction</i>
|
|
</que>
|
|
<que>
|
|
<i f="none">a bad reconstruction</i>
|
|
</que>
|
|
</split>
|
|
|
|
</frame>
|
|
</ignore>
|
|
|
|
<frame title="Can you classify with it?">
|
|
|
|
<split>
|
|
<que>
|
|
<list>
|
|
<e>Here compared to QCDorWhat</e>
|
|
<e>not really great, but not that bad either</e>
|
|
<e>and remember:only 4/6 particles were used</e>
|
|
</list>
|
|
</que>
|
|
<que>
|
|
<i f="lowcompare"></i>
|
|
</que>
|
|
</split>
|
|
|
|
</frame>
|
|
|
|
<ignore>
|
|
|
|
<split>
|
|
<que>
|
|
<list>
|
|
<e></e>
|
|
<e></e>
|
|
<e></e>
|
|
</list>
|
|
</que>
|
|
<que>
|
|
|
|
</que>
|
|
</split>
|
|
|
|
<frame>
|
|
|
|
<split>
|
|
<que>
|
|
<list>
|
|
<e></e>
|
|
<e></e>
|
|
<e></e>
|
|
</list>
|
|
</que>
|
|
<que>
|
|
<i f="none"></i>
|
|
</que>
|
|
</split>
|
|
|
|
</frame>
|
|
|
|
<frame>
|
|
|
|
<list>
|
|
<e></e>
|
|
<e></e>
|
|
<e></e>
|
|
</list>
|
|
|
|
</frame>
|
|
|
|
|
|
</ignore>
|
|
|
|
|
|
|