parsing two different strings produces identical ast.Str nodes:

import ast

txt1 = '"""\\n"""'
txt2 = '"""\n"""'

tree1 = ast.parse(txt1)
tree2 = ast.parse(txt2)

print(tree1.body[0].value.s == tree2.body[0].value.s)
print(bytes(tree1.body[0].value.s, encoding='utf-8'))
print(bytes(tree2.body[0].value.s, encoding='utf-8'))

>>> True
>>> b'\n'
>>> b'\n'

Expected result: I should be able to distinguish between the nodes created from two different strings.

Same problem holds for tabs (\\t vs \t).

For \\r vs \r, it is even more fun: 
txt1 = '"""\\r"""'
txt2 = '"""\r"""'

>>> b'\r'
>>> b'\n'

There was a recent discussion about this on the bug tracker, but of course now I can't find it. But for an earlier discussion, see issue 25885.

The decision was that your expectation that the nodes be distinguishable isn't a design goal of the ast. There's no expectation that code can round-trip through the ast and produce the original text, even for a string.