diff --git a/gpytorch/likelihoods/bernoulli_likelihood.py b/gpytorch/likelihoods/bernoulli_likelihood.py
index 614dc6546..62b9466fb 100644
--- a/gpytorch/likelihoods/bernoulli_likelihood.py
+++ b/gpytorch/likelihoods/bernoulli_likelihood.py
@@ -47,7 +47,7 @@ def log_marginal(
         return marginal.log_prob(observations)
 
     def marginal(self, function_dist: MultivariateNormal, *args: Any, **kwargs: Any) -> Bernoulli:
-        """
+        r"""
         :return: Analytic marginal :math:`p(\mathbf y)`.
         """
         mean = function_dist.mean
diff --git a/gpytorch/likelihoods/gaussian_likelihood.py b/gpytorch/likelihoods/gaussian_likelihood.py
index 5870a4e64..42cd715e1 100644
--- a/gpytorch/likelihoods/gaussian_likelihood.py
+++ b/gpytorch/likelihoods/gaussian_likelihood.py
@@ -130,7 +130,7 @@ def raw_noise(self, value: Tensor) -> None:
         self.noise_covar.initialize(raw_noise=value)
 
     def marginal(self, function_dist: MultivariateNormal, *args: Any, **kwargs: Any) -> MultivariateNormal:
-        """
+        r"""
         :return: Analytic marginal :math:`p(\mathbf y)`.
         """
         return super().marginal(function_dist, *args, **kwargs)
@@ -186,7 +186,7 @@ def log_marginal(
         return res * ~missing_idx
 
     def marginal(self, function_dist: MultivariateNormal, *args: Any, **kwargs: Any) -> MultivariateNormal:
-        """
+        r"""
         :return: Analytic marginal :math:`p(\mathbf y)`.
         """
         return super().marginal(function_dist, *args, **kwargs)
@@ -306,14 +306,14 @@ def _shaped_noise_covar(self, base_shape: torch.Size, *params: Any, **kwargs: An
         return res
 
     def marginal(self, function_dist: MultivariateNormal, *args: Any, **kwargs: Any) -> MultivariateNormal:
-        """
+        r"""
         :return: Analytic marginal :math:`p(\mathbf y)`.
         """
         return super().marginal(function_dist, *args, **kwargs)
 
 
 class DirichletClassificationLikelihood(FixedNoiseGaussianLikelihood):
-    """
+    r"""
     A classification likelihood that treats the labels as regression targets with fixed heteroscedastic noise.
     From Milios et al, NeurIPS, 2018 [https://arxiv.org/abs/1805.10915].
 
@@ -408,7 +408,7 @@ def get_fantasy_likelihood(self, **kwargs: Any) -> "DirichletClassificationLikel
         return fantasy_liklihood
 
     def marginal(self, function_dist: MultivariateNormal, *args: Any, **kwargs: Any) -> MultivariateNormal:
-        """
+        r"""
         :return: Analytic marginal :math:`p(\mathbf y)`.
         """
         return super().marginal(function_dist, *args, **kwargs)
diff --git a/gpytorch/likelihoods/multitask_gaussian_likelihood.py b/gpytorch/likelihoods/multitask_gaussian_likelihood.py
index 6864ab801..06f6e06c7 100644
--- a/gpytorch/likelihoods/multitask_gaussian_likelihood.py
+++ b/gpytorch/likelihoods/multitask_gaussian_likelihood.py
@@ -293,7 +293,7 @@ def _eval_covar_matrix(self) -> Tensor:
     def marginal(
         self, function_dist: MultitaskMultivariateNormal, *args: Any, **kwargs: Any
     ) -> MultitaskMultivariateNormal:
-        """
+        r"""
         :return: Analytic marginal :math:`p(\mathbf y)`.
         """
         return super().marginal(function_dist, *args, **kwargs)