diff --git a/src/sampler/mod.rs b/src/sampler/mod.rs
index 0f0b600..487b7f4 100644
--- a/src/sampler/mod.rs
+++ b/src/sampler/mod.rs
@@ -1,16 +1,11 @@
 //! Sampler trait and implementations for parameter sampling.
 
 pub mod grid;
-pub mod multivariate_tpe;
 pub mod random;
 pub mod tpe;
 
 use std::collections::HashMap;
 
-pub use multivariate_tpe::{
-    ConstantLiarStrategy, MultivariateTpeSampler, MultivariateTpeSamplerBuilder,
-};
-
 use crate::distribution::Distribution;
 use crate::param::ParamValue;
 
diff --git a/src/sampler/tpe/mod.rs b/src/sampler/tpe/mod.rs
index 407fa9b..3013111 100644
--- a/src/sampler/tpe/mod.rs
+++ b/src/sampler/tpe/mod.rs
@@ -4,9 +4,13 @@
 //! including support for intersection search space calculation.
 
 mod gamma;
+mod multivariate;
 mod sampler;
 pub mod search_space;
 
 pub use gamma::{FixedGamma, GammaStrategy, HyperoptGamma, LinearGamma, SqrtGamma};
+pub use multivariate::{
+    ConstantLiarStrategy, MultivariateTpeSampler, MultivariateTpeSamplerBuilder,
+};
 pub use sampler::{TpeSampler, TpeSamplerBuilder};
 pub use search_space::{GroupDecomposedSearchSpace, IntersectionSearchSpace};
diff --git a/src/sampler/multivariate_tpe.rs b/src/sampler/tpe/multivariate.rs
similarity index 99%
rename from src/sampler/multivariate_tpe.rs
rename to src/sampler/tpe/multivariate.rs
index e040eb2..7967222 100644
--- a/src/sampler/multivariate_tpe.rs
+++ b/src/sampler/tpe/multivariate.rs
@@ -1,6 +1,6 @@
 //! Multivariate Tree-Parzen Estimator (TPE) sampler implementation.
 //!
-//! Unlike the standard [`TpeSampler`](super::tpe::TpeSampler) which samples parameters
+//! Unlike the standard [`TpeSampler`] which samples parameters
 //! independently, the `MultivariateTpeSampler` models joint distributions over multiple
 //! parameters. This allows it to capture correlations between parameters, which can
 //! significantly improve optimization performance on problems where parameters interact.
@@ -12,7 +12,7 @@
 //! - The optimal value of one parameter depends on another
 //! - You have a fixed search space across all trials
 //!
-//! Use the standard [`TpeSampler`](super::tpe::TpeSampler) when:
+//! Use the standard [`TpeSampler`] when:
 //! - Parameters are independent
 //! - The search space varies dynamically between trials
 //! - You want simpler, faster optimization
@@ -68,7 +68,7 @@
 //! ## Basic Usage
 //!
 //! ```
-//! use optimizer::sampler::MultivariateTpeSampler;
+//! use optimizer::sampler::tpe::MultivariateTpeSampler;
 //!
 //! let sampler = MultivariateTpeSampler::builder()
 //!     .gamma(0.15)
@@ -82,8 +82,7 @@
 //! ## With Custom Gamma Strategy and Group Decomposition
 //!
 //! ```
-//! use optimizer::sampler::MultivariateTpeSampler;
-//! use optimizer::sampler::tpe::SqrtGamma;
+//! use optimizer::sampler::tpe::{MultivariateTpeSampler, SqrtGamma};
 //!
 //! let sampler = MultivariateTpeSampler::builder()
 //!     .gamma_strategy(SqrtGamma::default())
@@ -95,7 +94,7 @@
 //! ## Parallel Optimization with Constant Liar
 //!
 //! ```
-//! use optimizer::sampler::{ConstantLiarStrategy, MultivariateTpeSampler};
+//! use optimizer::sampler::tpe::{ConstantLiarStrategy, MultivariateTpeSampler};
 //!
 //! // Use mean imputation for parallel workers
 //! let sampler = MultivariateTpeSampler::builder()
@@ -108,7 +107,7 @@
 //! ## Suppressing Fallback Warnings
 //!
 //! ```
-//! use optimizer::sampler::MultivariateTpeSampler;
+//! use optimizer::sampler::tpe::MultivariateTpeSampler;
 //!
 //! let sampler = MultivariateTpeSampler::builder().build().unwrap();
 //! ```
@@ -120,11 +119,11 @@ use parking_lot::Mutex;
 use rand::SeedableRng;
 use rand::rngs::StdRng;
 
-use super::tpe::{FixedGamma, GammaStrategy};
-use super::{CompletedTrial, PendingTrial, Sampler};
+use super::{FixedGamma, GammaStrategy};
 use crate::distribution::Distribution;
 use crate::error::Result;
 use crate::param::ParamValue;
+use crate::sampler::{CompletedTrial, PendingTrial, Sampler};
 
 /// Strategy for imputing objective values for pending/running trials during parallel optimization.
 ///
@@ -143,7 +142,7 @@ use crate::param::ParamValue;
 /// # Examples
 ///
 /// ```
-/// use optimizer::sampler::ConstantLiarStrategy;
+/// use optimizer::sampler::tpe::ConstantLiarStrategy;
 ///
 /// // Use mean imputation for parallel optimization
 /// let strategy = ConstantLiarStrategy::Mean;
@@ -207,7 +206,7 @@ impl MultivariateTpeSampler {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSampler;
+    /// use optimizer::sampler::tpe::MultivariateTpeSampler;
     ///
     /// let sampler = MultivariateTpeSampler::new();
     /// ```
@@ -229,7 +228,7 @@ impl MultivariateTpeSampler {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSampler;
+    /// use optimizer::sampler::tpe::MultivariateTpeSampler;
     ///
     /// let sampler = MultivariateTpeSampler::builder()
     ///     .gamma(0.15)
@@ -446,7 +445,7 @@ impl MultivariateTpeSampler {
     ///
     /// ```ignore
     /// use std::collections::HashMap;
-    /// use optimizer::sampler::MultivariateTpeSampler;
+    /// use optimizer::sampler::tpe::MultivariateTpeSampler;
     /// use optimizer::sampler::tpe::IntersectionSearchSpace;
     ///
     /// let sampler = MultivariateTpeSampler::new();
@@ -473,10 +472,6 @@ impl MultivariateTpeSampler {
 
     /// Splits filtered trials into good and bad groups based on the gamma quantile.
     ///
-    /// This method is similar to [`TpeSampler::split_trials`](super::tpe::TpeSampler)
-    /// but operates on pre-filtered trials that have been selected for multivariate
-    /// sampling.
-    ///
     /// The gamma value is computed dynamically using the configured [`GammaStrategy`].
     /// Trials are sorted by objective value (ascending for minimization), and the
     /// gamma quantile determines the split point.
@@ -499,7 +494,7 @@ impl MultivariateTpeSampler {
     ///
     /// ```ignore
     /// use std::collections::HashMap;
-    /// use optimizer::sampler::MultivariateTpeSampler;
+    /// use optimizer::sampler::tpe::MultivariateTpeSampler;
     /// use optimizer::sampler::tpe::IntersectionSearchSpace;
     ///
     /// let sampler = MultivariateTpeSampler::new();
@@ -593,7 +588,7 @@ impl MultivariateTpeSampler {
     ///
     /// ```ignore
     /// use std::collections::HashMap;
-    /// use optimizer::sampler::MultivariateTpeSampler;
+    /// use optimizer::sampler::tpe::MultivariateTpeSampler;
     ///
     /// let sampler = MultivariateTpeSampler::new();
     /// let trials = vec![/* ... completed trials ... */];
@@ -661,7 +656,7 @@ impl MultivariateTpeSampler {
     /// # Examples
     ///
     /// ```ignore
-    /// use optimizer::sampler::MultivariateTpeSampler;
+    /// use optimizer::sampler::tpe::MultivariateTpeSampler;
     /// use optimizer::kde::MultivariateKDE;
     ///
     /// let sampler = MultivariateTpeSampler::builder()
@@ -859,7 +854,7 @@ impl MultivariateTpeSampler {
     ///
     /// ```ignore
     /// use std::collections::HashMap;
-    /// use optimizer::sampler::MultivariateTpeSampler;
+    /// use optimizer::sampler::tpe::MultivariateTpeSampler;
     /// use optimizer::distribution::{Distribution, FloatDistribution};
     ///
     /// let sampler = MultivariateTpeSampler::builder()
@@ -925,7 +920,7 @@ impl MultivariateTpeSampler {
     ) -> HashMap<String, ParamValue> {
         use std::collections::HashSet;
 
-        use crate::sampler::tpe::GroupDecomposedSearchSpace;
+        use super::GroupDecomposedSearchSpace;
 
         // Decompose the search space into independent parameter groups
         let groups = GroupDecomposedSearchSpace::calculate(history);
@@ -1014,8 +1009,8 @@ impl MultivariateTpeSampler {
         history: &[CompletedTrial],
         rng: &mut StdRng,
     ) -> HashMap<String, ParamValue> {
+        use super::IntersectionSearchSpace;
         use crate::kde::MultivariateKDE;
-        use crate::sampler::tpe::IntersectionSearchSpace;
 
         // Early returns for cases requiring random sampling
         if history.len() < self.n_startup_trials {
@@ -1172,7 +1167,7 @@ impl MultivariateTpeSampler {
     ///
     /// This method is used to sample parameters that are not in the intersection
     /// search space. It uses independent univariate TPE sampling for each parameter,
-    /// similar to the standard [`TpeSampler`](super::tpe::TpeSampler).
+    /// similar to the standard [`TpeSampler`].
     ///
     /// When there isn't enough history for a parameter, falls back to uniform sampling.
     #[allow(dead_code)]
@@ -1772,7 +1767,7 @@ impl MultivariateTpeSampler {
 /// Using a fixed gamma value:
 ///
 /// ```
-/// use optimizer::sampler::MultivariateTpeSamplerBuilder;
+/// use optimizer::sampler::tpe::MultivariateTpeSamplerBuilder;
 ///
 /// let sampler = MultivariateTpeSamplerBuilder::new()
 ///     .gamma(0.15)
@@ -1786,8 +1781,7 @@ impl MultivariateTpeSampler {
 /// Using a custom gamma strategy:
 ///
 /// ```
-/// use optimizer::sampler::MultivariateTpeSamplerBuilder;
-/// use optimizer::sampler::tpe::SqrtGamma;
+/// use optimizer::sampler::tpe::{MultivariateTpeSamplerBuilder, SqrtGamma};
 ///
 /// let sampler = MultivariateTpeSamplerBuilder::new()
 ///     .gamma_strategy(SqrtGamma::default())
@@ -1845,7 +1839,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSamplerBuilder;
+    /// use optimizer::sampler::tpe::MultivariateTpeSamplerBuilder;
     ///
     /// let sampler = MultivariateTpeSamplerBuilder::new()
     ///     .gamma(0.10)  // Use top 10% as "good" trials
@@ -1876,8 +1870,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSamplerBuilder;
-    /// use optimizer::sampler::tpe::{LinearGamma, SqrtGamma};
+    /// use optimizer::sampler::tpe::{LinearGamma, MultivariateTpeSamplerBuilder, SqrtGamma};
     ///
     /// // Square root strategy (Optuna-style)
     /// let sampler = MultivariateTpeSamplerBuilder::new()
@@ -1911,7 +1904,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSamplerBuilder;
+    /// use optimizer::sampler::tpe::MultivariateTpeSamplerBuilder;
     ///
     /// let sampler = MultivariateTpeSamplerBuilder::new()
     ///     .n_startup_trials(20)  // Random sample first 20 trials
@@ -1937,7 +1930,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSamplerBuilder;
+    /// use optimizer::sampler::tpe::MultivariateTpeSamplerBuilder;
     ///
     /// let sampler = MultivariateTpeSamplerBuilder::new()
     ///     .n_ei_candidates(48)  // Evaluate more candidates
@@ -1964,7 +1957,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSamplerBuilder;
+    /// use optimizer::sampler::tpe::MultivariateTpeSamplerBuilder;
     ///
     /// let sampler = MultivariateTpeSamplerBuilder::new()
     ///     .group(true)  // Enable group decomposition
@@ -1990,7 +1983,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::{ConstantLiarStrategy, MultivariateTpeSamplerBuilder};
+    /// use optimizer::sampler::tpe::{ConstantLiarStrategy, MultivariateTpeSamplerBuilder};
     ///
     /// // Use mean imputation for pending trials
     /// let sampler = MultivariateTpeSamplerBuilder::new()
@@ -2025,7 +2018,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSamplerBuilder;
+    /// use optimizer::sampler::tpe::MultivariateTpeSamplerBuilder;
     ///
     /// let sampler = MultivariateTpeSamplerBuilder::new()
     ///     .seed(42)  // Reproducible results
@@ -2047,7 +2040,7 @@ impl MultivariateTpeSamplerBuilder {
     /// # Examples
     ///
     /// ```
-    /// use optimizer::sampler::MultivariateTpeSamplerBuilder;
+    /// use optimizer::sampler::tpe::MultivariateTpeSamplerBuilder;
     ///
     /// let sampler = MultivariateTpeSamplerBuilder::new()
     ///     .gamma(0.15)
diff --git a/src/sampler/tpe/sampler.rs b/src/sampler/tpe/sampler.rs
index 59e1e8c..434967a 100644
--- a/src/sampler/tpe/sampler.rs
+++ b/src/sampler/tpe/sampler.rs
@@ -280,6 +280,7 @@ impl TpeSampler {
         clippy::cast_possible_truncation,
         clippy::cast_sign_loss
     )]
+    #[must_use]
     fn split_trials<'a>(
         &self,
         history: &'a [CompletedTrial],
diff --git a/tests/multivariate_tpe_integration.rs b/tests/multivariate_tpe_integration.rs
index c2c6689..21b54f9 100644
--- a/tests/multivariate_tpe_integration.rs
+++ b/tests/multivariate_tpe_integration.rs
@@ -9,8 +9,7 @@
     clippy::cast_possible_truncation
 )]
 
-use optimizer::sampler::MultivariateTpeSampler;
-use optimizer::sampler::tpe::TpeSampler;
+use optimizer::sampler::tpe::{MultivariateTpeSampler, TpeSampler};
 use optimizer::{Direction, Error, Study};
 
 // =============================================================================