add more clarity on spec and align with real implementation

docs/ops/activation/SoftPlus_4.md

@@ -19,16 +19,19 @@ SoftPlus(x) = \left\{\begin{array}{r}
 \end{array}\right.
 \f]
 
-**Note**: For numerical stability the operation reverts to the linear function when `x > threshold` where `threshold` depends on *T* and
-is chosen in such a way that the difference between the linear function and exact calculation is no more than `1e-6`.
-The `threshold` can be calculated with the following formula where `alpha` is the number of digits after the decimal point,
-`beta` is maximum value of *T* data type:
+**Note**: 
+
+For numerical stability the operation reverts to the linear function when `x > threshold` where `threshold` depends on *T* and is chosen in such a way that the difference between the linear function and exact calculation is no more than `1e-6`.
+
+The `threshold` is internally defined parameter, which can be calculated with the following formula where `alpha` is the number of digits after the decimal point, `beta` is maximum value of *T* data type:
 
 \f[
 -log(e^{10^{-\alpha}} - 1.0) < threshold < log(\beta)
 \f]
 
-For example, if *T* is `fp32`, `threshold` should be `20` or if *T* is `fp16`, `threshold` should be `11`.
+For example:
+If *T* is `fp32`, `threshold` range is `13.8 ~ 88.7`, in practice `threshold` is set to be `20`.
+If *T* is `fp16`, in practice `threshold` is also set to be `20`，which is also ok as the data is operated in `fp32` format and then converted to `fp16` at the end.
 
 **Attributes**: *SoftPlus* operation has no attributes.