3 ms·
Worth opening the pdf just for the graph on page 1.
by andai 1y ago
Worth opening the pdf just for the graph on page 1.
- mNovak 1y agoReally seems like the reason logarithmic scales were invented..
- arscan 1y agoIn case you want to know what’s going on in the left side of that chart, they gave a log scale in appendix a. I was thinking it was silly to not just use that version on the top, but I guess log scales make big differences ’feel’ smaller.
- kevmo314 1y agoAt that rate, they might as well have gone one step further and made the x axis exponential scale to make it feel even bigger.
- levocardia 1y agoA log scale is actually appropriate in this context from a first-principles perspective. Per scaling laws (and also general behavior of epsilon-probability of failure multiplied N times), you would generally expect more vs. less effective techniques to have multiplicatively greater or fewer steps until failure, not additively greater/fewer. Figure 1 is comical, but the appendix figure is the more scientifically appropriate one.
- esafak 1y agoA striking example of how not to present data. If the Cognizant AI team is here: please can you fix it in the next version of the paper?
- rdlw 1y agoI think it's a brilliant example of how to use data to make a point. https://xkcd.com/1162/ https://xkcd.com/1162/
- datadrivenangel 1y agohttp://www.vibechart.net http://www.vibechart.net
- mrbungie 1y agoObviously, not the best plot to use according to Data Visualization theory and common practice, but I think it candidly conveys the point anyway. As someone else points, the data is the worrying aspect, as it points towards state-of-the-art models not being able of making more than 0 consecutive steps without errors.
- vessenes 1y agoI was just thinking "these guys will talk about this graph for the rest of their lives", it's the best graph you could ever hope to put into a paper. Loved it.
- causalmodels 1y agoThe dashed lines on top of the data points and labels is making me wince