diff --git a/test/utils.jl b/test/utils.jl
index 2cef162fe..0d66e9f67 100644
--- a/test/utils.jl
+++ b/test/utils.jl
@@ -6,6 +6,10 @@
     v = randn(rng, D)
     w = randn(rng, N)
 
+    @testset "check_args macro" begin
+        @test_throws ArgumentError GammaExponentialKernel(-1.0, Euclidean())
+    end
+
     @testset "VecOfVecs" begin
         @test vec_of_vecs(X; obsdim=2) == ColVecs(X)
         @test vec_of_vecs(X; obsdim=1) == RowVecs(X)
@@ -42,6 +46,10 @@
         KernelFunctions.pairwise!(K, SqEuclidean(), DX, DY)
         @test K ≈ pairwise(SqEuclidean(), X, Y; dims=2)
 
+        y = rand(N, 1)
+        yv = y[:]
+        @test RowVecs(y) == RowVecs(yv)
+
         let
             @test Zygote.pullback(ColVecs, X)[1] == DX
             DX, back = Zygote.pullback(ColVecs, X)
@@ -98,14 +106,40 @@
         x_rowvecs = RowVecs(randn(7, 3))
 
         @test isapprox(
-            pairwise(SqEuclidean(), x_colvecs, x_rowvecs),
+            KernelFunctions.pairwise(SqEuclidean(), x_colvecs, x_rowvecs),
             pairwise(SqEuclidean(), collect(x_colvecs), collect(x_rowvecs)),
         )
         @test isapprox(
-            pairwise(SqEuclidean(), x_rowvecs, x_colvecs),
+            KernelFunctions.pairwise(SqEuclidean(), x_rowvecs, x_colvecs),
             pairwise(SqEuclidean(), collect(x_rowvecs), collect(x_colvecs)),
         )
     end
+    @testset "AbstractVector + RowVecs" begin
+        x = [randn(3) for _ in 1:5]
+        x_rowvecs = RowVecs(randn(7, 3))
+
+        @test isapprox(
+            KernelFunctions.pairwise(SqEuclidean(), x, x_rowvecs),
+            pairwise(SqEuclidean(), x, collect(x_rowvecs)),
+        )
+        @test isapprox(
+            KernelFunctions.pairwise(SqEuclidean(), x_rowvecs, x),
+            pairwise(SqEuclidean(), collect(x_rowvecs), x),
+        )
+    end
+    @testset "AbstractVector + ColVecs" begin
+        x = [randn(3) for _ in 1:5]
+        x_colvecs = ColVecs(randn(3, 7))
+
+        @test isapprox(
+            KernelFunctions.pairwise(SqEuclidean(), x, x_colvecs),
+            pairwise(SqEuclidean(), x, collect(x_colvecs)),
+        )
+        @test isapprox(
+            KernelFunctions.pairwise(SqEuclidean(), x_colvecs, x),
+            pairwise(SqEuclidean(), collect(x_colvecs), x),
+        )
+    end
     @testset "input checks" begin
         D = 3
         D⁻ = 2