diff --git a/cpp/benchmarks/stream_compaction/distinct.cpp b/cpp/benchmarks/stream_compaction/distinct.cpp
index d7deebca89a..75d04bb4e8e 100644
--- a/cpp/benchmarks/stream_compaction/distinct.cpp
+++ b/cpp/benchmarks/stream_compaction/distinct.cpp
@@ -34,6 +34,7 @@ void nvbench_distinct(nvbench::state& state, nvbench::type_list<Type>)
   cudf::size_type const num_rows    = state.get_int64("NumRows");
   auto const keep                   = get_keep(state.get_string("keep"));
   cudf::size_type const cardinality = state.get_int64("cardinality");
+  auto const null_probability       = state.get_float64("null_probability");
 
   if (cardinality > num_rows) {
     state.skip("cardinality > num_rows");
@@ -42,7 +43,7 @@ void nvbench_distinct(nvbench::state& state, nvbench::type_list<Type>)
 
   data_profile profile = data_profile_builder()
                            .cardinality(cardinality)
-                           .null_probability(0.01)
+                           .null_probability(null_probability)
                            .distribution(cudf::type_to_id<Type>(),
                                          distribution_id::UNIFORM,
                                          static_cast<Type>(0),
@@ -65,6 +66,7 @@ using data_type = nvbench::type_list<int32_t, int64_t>;
 NVBENCH_BENCH_TYPES(nvbench_distinct, NVBENCH_TYPE_AXES(data_type))
   .set_name("distinct")
   .set_type_axes_names({"Type"})
+  .add_float64_axis("null_probability", {0.01})
   .add_string_axis("keep", {"any", "first", "last", "none"})
   .add_int64_axis("cardinality", {100, 100'000, 10'000'000, 1'000'000'000})
   .add_int64_axis("NumRows", {100, 100'000, 10'000'000, 1'000'000'000});
diff --git a/cpp/src/stream_compaction/distinct.cu b/cpp/src/stream_compaction/distinct.cu
index 7d11b02d3e1..9ab8ed5938a 100644
--- a/cpp/src/stream_compaction/distinct.cu
+++ b/cpp/src/stream_compaction/distinct.cu
@@ -95,8 +95,8 @@ rmm::device_uvector<size_type> distinct_indices(table_view const& input,
   auto const row_equal = cudf::experimental::row::equality::self_comparator(preprocessed_input);
 
   auto const helper_func = [&](auto const& d_equal) {
-    using RowHasher = std::decay_t<decltype(d_equal)>;
-    auto set        = hash_set_type<RowHasher>{
+    using RowEqual = std::decay_t<decltype(d_equal)>;
+    auto set       = distinct_set_t<RowEqual>{
       num_rows,
       0.5,  // desired load factor
       cuco::empty_key{cudf::detail::CUDF_SIZE_TYPE_SENTINEL},
diff --git a/cpp/src/stream_compaction/distinct_helpers.cu b/cpp/src/stream_compaction/distinct_helpers.cu
index c3a004b7f28..aadb438b019 100644
--- a/cpp/src/stream_compaction/distinct_helpers.cu
+++ b/cpp/src/stream_compaction/distinct_helpers.cu
@@ -21,8 +21,8 @@
 
 namespace cudf::detail {
 
-template <typename RowHasher>
-rmm::device_uvector<size_type> reduce_by_row(hash_set_type<RowHasher>& set,
+template <typename RowEqual>
+rmm::device_uvector<size_type> reduce_by_row(distinct_set_t<RowEqual>& set,
                                              size_type num_rows,
                                              duplicate_keep_option keep,
                                              rmm::cuda_stream_view stream,
@@ -100,7 +100,7 @@ rmm::device_uvector<size_type> reduce_by_row(hash_set_type<RowHasher>& set,
 }
 
 template rmm::device_uvector<size_type> reduce_by_row(
-  hash_set_type<cudf::experimental::row::equality::device_row_comparator<
+  distinct_set_t<cudf::experimental::row::equality::device_row_comparator<
     false,
     cudf::nullate::DYNAMIC,
     cudf::experimental::row::equality::nan_equal_physical_equality_comparator>>& set,
@@ -110,7 +110,7 @@ template rmm::device_uvector<size_type> reduce_by_row(
   rmm::device_async_resource_ref mr);
 
 template rmm::device_uvector<size_type> reduce_by_row(
-  hash_set_type<cudf::experimental::row::equality::device_row_comparator<
+  distinct_set_t<cudf::experimental::row::equality::device_row_comparator<
     true,
     cudf::nullate::DYNAMIC,
     cudf::experimental::row::equality::nan_equal_physical_equality_comparator>>& set,
@@ -120,7 +120,7 @@ template rmm::device_uvector<size_type> reduce_by_row(
   rmm::device_async_resource_ref mr);
 
 template rmm::device_uvector<size_type> reduce_by_row(
-  hash_set_type<cudf::experimental::row::equality::device_row_comparator<
+  distinct_set_t<cudf::experimental::row::equality::device_row_comparator<
     false,
     cudf::nullate::DYNAMIC,
     cudf::experimental::row::equality::physical_equality_comparator>>& set,
@@ -130,7 +130,7 @@ template rmm::device_uvector<size_type> reduce_by_row(
   rmm::device_async_resource_ref mr);
 
 template rmm::device_uvector<size_type> reduce_by_row(
-  hash_set_type<cudf::experimental::row::equality::device_row_comparator<
+  distinct_set_t<cudf::experimental::row::equality::device_row_comparator<
     true,
     cudf::nullate::DYNAMIC,
     cudf::experimental::row::equality::physical_equality_comparator>>& set,
diff --git a/cpp/src/stream_compaction/distinct_helpers.hpp b/cpp/src/stream_compaction/distinct_helpers.hpp
index f15807c2434..4ca1cab937a 100644
--- a/cpp/src/stream_compaction/distinct_helpers.hpp
+++ b/cpp/src/stream_compaction/distinct_helpers.hpp
@@ -47,12 +47,12 @@ auto constexpr reduction_init_value(duplicate_keep_option keep)
   }
 }
 
-template <typename RowHasher>
-using hash_set_type =
+template <typename RowEqual>
+using distinct_set_t =
   cuco::static_set<size_type,
                    cuco::extent<int64_t>,
                    cuda::thread_scope_device,
-                   RowHasher,
+                   RowEqual,
                    cuco::linear_probing<1,
                                         cudf::experimental::row::hash::device_row_hasher<
                                           cudf::hashing::detail::default_hash,
@@ -79,6 +79,8 @@ using hash_set_type =
  * the `reduction_init_value()` function. Then, the reduction result for each row group is written
  * into the output array at the index of an unspecified row in the group.
  *
+ * @tparam RowEqual The type of row equality comparator
+ *
  * @param set The auxiliary set to perform reduction
  * @param set_size The number of elements in set
  * @param num_rows The number of all input rows
@@ -87,8 +89,8 @@ using hash_set_type =
  * @param mr Device memory resource used to allocate the returned vector
  * @return A device_uvector containing the output indices
  */
-template <typename RowHasher>
-rmm::device_uvector<size_type> reduce_by_row(hash_set_type<RowHasher>& set,
+template <typename RowEqual>
+rmm::device_uvector<size_type> reduce_by_row(distinct_set_t<RowEqual>& set,
                                              size_type num_rows,
                                              duplicate_keep_option keep,
                                              rmm::cuda_stream_view stream,