markdown fix

Author: Hyeok Kim

static/blog-assets/posts/2019-08-12-errudite.md

@@ -107,7 +107,7 @@ If we apply all the partial filters and build all the other groups, we notice a
 So, the DSL helps cover the entire dataset, including the correct instances. The error analysis is more systematic and scalable this way, and can give you different conclusions when compared to looking at a small sample of mistakes only. We formally state our second principle as:
 
 <blockquote>
-**P2: Error prevalence should be assessed over the entire dataset — including the true positive (non-error) examples.**
+<em>P2: Error prevalence should be assessed over the entire dataset — including the true positive (non-error) examples.</em>
 </blockquote>
 
 * * *
@@ -121,7 +121,7 @@ Now, we’ve defined groups related to distractors. But, the presence of distrac
 This leads to one more problem in the status-quo: *We cannot effectively isolate the true cause of an error.* To dig into root causes, we state a third principle:
 
 <blockquote>
-**P3: Error hypotheses should be explicitly tested.**
+<em>P3: Error hypotheses should be explicitly tested.</em>
 </blockquote>
 
 In Errudite, we help answer this question, *“Are the 192 instances really wrong because of the distractor?”*, by asking a related what-if question: *“If the predicted distractor was not there, would the model predict correctly?”* We answer this question though counterfactual analysis with rewrite rules.