diff --git a/mlir/unittests/Dialect/SparseTensor/CMakeLists.txt b/mlir/unittests/Dialect/SparseTensor/CMakeLists.txt
new file mode 100644
--- /dev/null
+++ b/mlir/unittests/Dialect/SparseTensor/CMakeLists.txt
@@ -0,0 +1,7 @@
+add_mlir_unittest(MLIRSparseTensorTests
+  MergerTest.cpp
+)
+target_link_libraries(MLIRSparseTensorTests
+  PRIVATE
+  MLIRSparseTensorUtils
+)
diff --git a/mlir/unittests/Dialect/SparseTensor/MergerTest.cpp b/mlir/unittests/Dialect/SparseTensor/MergerTest.cpp
new file mode 100644
--- /dev/null
+++ b/mlir/unittests/Dialect/SparseTensor/MergerTest.cpp
@@ -0,0 +1,166 @@
+#include "mlir/Dialect/SparseTensor/Utils/Merger.h"
+#include "gmock/gmock.h"
+#include "gtest/gtest.h"
+
+using namespace mlir;
+using namespace mlir::sparse_tensor;
+
+namespace {
+
+const unsigned NUM_TENSORS = 2, NUM_LOOPS = 1;
+
+class MergerTest : public ::testing::Test {
+protected:
+  MergerTest() : merger(NUM_TENSORS, NUM_LOOPS) {
+
+    // Tensor 0: sparse vector.
+    merger.addExp(Kind::kTensor, 0, -1u);
+    merger.setDim(0, 0, Dim::kSparse);
+
+    // Tensor 1: sparse vector.
+    merger.addExp(Kind::kTensor, 1, -1u);
+    merger.setDim(1, 0, Dim::kSparse);
+  }
+
+  ///
+  /// Expression construction helpers.
+  ///
+
+  unsigned tensor(unsigned tensor) { return tensor; }
+
+  unsigned addf(unsigned e0, unsigned e1) {
+    return merger.addExp(Kind::kAddF, e0, e1);
+  }
+
+  unsigned mulf(unsigned e0, unsigned e1) {
+    return merger.addExp(Kind::kMulF, e0, e1);
+  }
+
+  ///
+  /// Comparison helpers.
+  ///
+
+  /// Check that the expression of lat point p in set s is equivalent to the
+  /// expression pointed to by e.
+  void expectLatPointExpressionEquals(unsigned s, unsigned p, unsigned e) {
+    EXPECT_TRUE(compareExpression(merger.lat(merger.set(s)[p]).exp, e));
+  }
+
+  /// Checks that the true bits of the lat point p in set s correspond to the
+  /// bits indicated by the (loop, tensor) pairs in loops.
+  void
+  expectLatPointBitsEqual(unsigned s, unsigned p,
+                          std::vector<std::tuple<unsigned, unsigned>> loops) {
+    llvm::BitVector testBits = llvm::BitVector(NUM_TENSORS + 1, false);
+    for (auto l : loops) {
+      auto loop = std::get<0>(l);
+      auto tensor = std::get<1>(l);
+      testBits.set(NUM_TENSORS * loop + tensor);
+    }
+    EXPECT_THAT(merger.lat(merger.set(s)[p]).bits, testBits);
+  }
+
+  /// Check that there are n lattice points in set s.
+  void expectNumLatPoints(unsigned s, unsigned n) {
+    EXPECT_THAT(merger.set(s).size(), n);
+  }
+
+  Merger merger;
+
+private:
+  /// Compares e0 and e1 for equality. Equality is defined recursively as:
+  /// - Two expressions can only be equal if they have the same Kind.
+  /// - Two binary expressions are equal if they have the same Kind and their
+  ///     children are equal.
+  /// - Expressions with Kind invariant or tensor are equal if they have the
+  ///     same expression id.
+  bool compareExpression(unsigned e0, unsigned e1) {
+    auto tensorExp0 = merger.exp(e0);
+    auto tensorExp1 = merger.exp(e1);
+    if (tensorExp0.kind != tensorExp1.kind) {
+      return false;
+    }
+    switch (tensorExp0.kind) {
+    case Kind::kAddF:
+    case Kind::kMulF:
+    case Kind::kAddI:
+    case Kind::kMulI:
+      return compareExpression(tensorExp0.e0, tensorExp1.e0) &&
+             compareExpression(tensorExp0.e1, tensorExp1.e1);
+    case Kind::kInvariant:
+      return tensorExp0.val == tensorExp1.val;
+    case Kind::kTensor:
+      return tensorExp0.e0 == tensorExp1.e0;
+    }
+  }
+};
+
+} // anonymous namespace
+
+/// Vector addition of 2 vectors, i.e.:
+///   a(i) = b(i) + c(i)
+/// which should form the 3 lattice points
+/// {
+///   lat( i_00 i_01 / (tensor_0 + tensor_1) )
+///   lat( i_00 / tensor_0 )
+///   lat( i_01 / tensor_1 )
+/// }
+/// and after optimization, will reduce to the 2 lattice points
+/// {
+///   lat( i_00 i_01 / (tensor_0 + tensor_1) )
+///   lat( i_00 / tensor_0 )
+/// }
+TEST_F(MergerTest, VectorAdd2) {
+  // Construct expression.
+  auto e = addf(tensor(0), tensor(1));
+
+  // Build lattices and check.
+  auto s = merger.buildLattices(e, 0);
+  expectNumLatPoints(s, 3);
+  expectLatPointExpressionEquals(s, /*lat point*/ 0,
+                                 addf(tensor(0), tensor(1)));
+  expectLatPointBitsEqual(
+      s, /*lat point*/ 0,
+      {{/*loop*/ 0, /*tensor*/ 0}, {/*loop*/ 0, /*tensor*/ 1}});
+  expectLatPointExpressionEquals(s, /*lat point*/ 1, tensor(0));
+  expectLatPointBitsEqual(s, /*lat point*/ 1, {{/*loop*/ 0, /*tensor*/ 0}});
+  expectLatPointExpressionEquals(s, /*lat point*/ 2, tensor(1));
+  expectLatPointBitsEqual(s, /*lat point*/ 2, {{/*loop*/ 0, /*tensor*/ 1}});
+
+  // Optimize lattices and check.
+  s = merger.optimizeSet(s);
+  expectNumLatPoints(s, 2);
+  expectLatPointExpressionEquals(s, 0, addf(tensor(0), tensor(1)));
+  expectLatPointBitsEqual(
+      s, /*lat point*/ 0,
+      {{/*loop*/ 0, /*tensor*/ 0}, {/*loop*/ 0, /*tensor*/ 1}});
+  expectLatPointExpressionEquals(s, 1, tensor(0));
+  expectLatPointBitsEqual(s, /*lat point*/ 1, {{/*loop*/ 0, /*tensor*/ 0}});
+}
+
+/// Vector multiplication of 2 vectors, i.e.:
+///   a(i) = b(i) * c(i)
+/// which should form the single lattice point
+/// {
+///   lat( i_00 i_01 / (tensor_0 * tensor_1) )
+/// }
+TEST_F(MergerTest, VectorMul2) {
+  // Construct expression.
+  auto e = mulf(tensor(0), tensor(1));
+
+  // Build lattices and check.
+  auto s = merger.buildLattices(e, 0);
+  expectNumLatPoints(s, 1);
+  expectLatPointExpressionEquals(s, 0, mulf(tensor(0), tensor(1)));
+  expectLatPointBitsEqual(
+      s, /*lat point*/ 0,
+      {{/*loop*/ 0, /*tensor*/ 0}, {/*loop*/ 0, /*tensor*/ 1}});
+
+  // Optimize lattices and check.
+  s = merger.optimizeSet(s);
+  expectNumLatPoints(s, 1);
+  expectLatPointExpressionEquals(s, 0, mulf(tensor(0), tensor(1)));
+  expectLatPointBitsEqual(
+      s, /*lat point*/ 0,
+      {{/*loop*/ 0, /*tensor*/ 0}, {/*loop*/ 0, /*tensor*/ 1}});
+}