Add softsign to tensorflow frontend api with test (ivy-llc#21352)

ivy/functional/frontends/tensorflow/math.py

@@ -805,3 +805,11 @@ def l2_normalize(x, axis=None, epsilon=1e-12, name=None):
     square_sum = ivy.sum(ivy.square(x), axis=axis, keepdims=True)
     x_inv_norm = ivy.reciprocal(ivy.sqrt(ivy.maximum(square_sum, epsilon)))
     return ivy.multiply(x, x_inv_norm)
+
+
+@with_supported_dtypes(
+    {"2.13.0 and below": ("bfloat32", "float32", "float64")}, "tensorflow"
+)
+@to_ivy_arrays_and_back
+def softsign(features, name=None):
+    return ivy.divide(features, ivy.abs(features) + 1)

ivy_tests/test_ivy/test_frontends/test_tensorflow/test_math.py

@@ -3171,3 +3171,32 @@ def test_tensorflow_cumprod(  # NOQA
         exclusive=exclusive,
         reverse=reverse,
     )
+
+
+# softsign
+@handle_frontend_test(
+    fn_tree="tensorflow.math.softsign",
+    dtype_and_x=helpers.dtype_and_values(
+        available_dtypes=helpers.get_dtypes("valid"),
+    ),
+    test_with_out=st.just(False),
+)
+def test_tensorflow_softsign(
+    *,
+    dtype_and_x,
+    on_device,
+    fn_tree,
+    frontend,
+    test_flags,
+    backend_fw,
+):
+    input_dtype, x = dtype_and_x
+    helpers.test_frontend_function(
+        input_dtypes=input_dtype,
+        backend_to_test=backend_fw,
+        frontend=frontend,
+        test_flags=test_flags,
+        fn_tree=fn_tree,
+        on_device=on_device,
+        features=x[0],
+    )