Update README.md

README.md

@@ -30,7 +30,7 @@ The paper reports experiments at **GPT-2** scale on **LM1B** and **OpenWebText**
 
 ## Method Summary
 
-For a clean token $x$, SCDD uses the following marginal forward distribution:
+For a clean token \\(x\\), SCDD uses the following marginal forward distribution:
 
 $$
 q(z_t \mid x) = \operatorname{Cat}\left(z_t;, \gamma_t\left(\rho_t x + (1-\rho_t)u\right) + (1-\gamma_t)m\right).
@@ -38,10 +38,10 @@ $$
 
 where:
 
-* `m` is the `[MASK]` token.
-* $`u`$ is the uniform distribution over non-`[MASK]` tokens.
-* $`\gamma_t`$ is the marginal probability that $z_t$ is not `[MASK]`.
-* $`\rho_t`$ is the marginal probability that $z_t$ retains the clean token among the non-`[MASK]` mass.
+* \\(m\\) is the `[MASK]` token.
+* \\(u\\) is the uniform distribution over non-`[MASK]` tokens.
+* \\(\gamma_t\\) is the marginal probability that \\(z_t\\) is not `[MASK]`.
+* \\(\rho_t\\) is the marginal probability that \\(z_t\\) retains the clean token among the non-`[MASK]` mass.