chore: make some documents clearer

src/int/add.rs

@@ -21,8 +21,8 @@ impl<const LIMBS: usize> Int<LIMBS> {
 
         // Step 2. determine whether overflow happened.
         // Note:
-        // - overflow can only happen when the inputs have the same sign, and then
-        // - overflow occurs if and only if the result has the opposite sign of both inputs.
+        // - overflow can only happen when the inputs have the same sign, and
+        // - overflow occurs if and only if the result has the opposite sign from both inputs.
         //
         // We can thus express the overflow flag as: (self.msb == rhs.msb) & (self.msb != res.msb)
         let self_msb = self.is_negative();

src/int/sign.rs

@@ -3,7 +3,7 @@ use num_traits::ConstZero;
 
 impl<const LIMBS: usize> Int<LIMBS> {
     /// Returns the word of most significant [`Limb`].
-    /// For the generative case where the number of limbs is zero,
+    /// For the degenerate case where the number of limbs is zero,
     /// zeroed word is returned (which is semantically correct).
     /// This method leaks the limb length of the value, which is also OK.
     const fn most_significant_word(&self) -> Word {

src/int/sub.rs

@@ -16,7 +16,7 @@ impl<const LIMBS: usize> Int<LIMBS> {
 
         // Step 2. check whether underflow happened.
         // Note:
-        // - underflow can only happen when the inputs have opposing signs, and then
+        // - underflow can only happen when the inputs have opposing signs, and
         // - underflow occurs if and only if the result and the lhs have opposing signs.
         //
         // We can thus express the overflow flag as: (self.msb != rhs.msb) & (self.msb != res.msb)