There's a couple of things worth considering here.
The first is that traditional/trope-laden Post-Apoc (PA for short) android fiction tends to rely heavily on a lot of hand-waving.
Nothing is created in a void, and the number of technologies, and specialisations which need to be drawn together to build a "super human" android are huge. And in a PA-world, a village would simply not have the ability to create their own technologies - everything from refining metals, creating alloys, machining parts, producing high-spec CPUs, etc require huge amounts of raw resources and vast amounts of knowledge, experience and specialisation. And that requires large numbers of people, which in turn require vast amounts of food, etc.
As such, if someone in a Post-Apoc (PA for short) world is able to build a high-spec humanoid android, then the only way they realistically can do that is by bolting together widely available and standardised existing technology.
Which means that our scientist is either bolting stuff together which was created pre-collapse, or has somehow traded with external entities to get the technologies he needs.
Which in turn means that if the rest of the world hasn't suffered a PA collapse, then any culture able to fund and sustain a "modern army" will be aware of this technology and will have been able to develop their own technologies and techniques to counter it. More importantly, they'll almost certainly have their own androids, which are likely to be more advanced than some junker thrown together in a PA village...
Equally, I'd question some of the features being given to this android. E.g. what materials is this android made of which can be immersed in liquid nitrogen (at least -340F/-196C) without becoming brittle[*]? Assuming the android looks human (given that you've given it a gender), then her skin and hair in particular will just snap and break off when it moves, so you'd end up with some sort of Terminator endo-skeleton wandering around. And even then, that's only if there's nothing within the android's body which would be equally prone to breaking if flexed at that temperature.
(Conversely, I guess the android could maintain an internal heat source, but a hell of a lot of power would be needed, it'd need to be able to deliver said heat to all of it's limb virtually instantaneously and all of that liquid nitrogen would immediately vapourise. Which would be a bit of a clue to watchers that there's something odd going on...
Either way, how many vats of liquid nitrogen will be left just lying around in a PA landscape?)
I could add more about the physics of some of the other things mentioned, especially when it comes to the "super intelligent" AI for this machine, but for now, I'll just flag the idea that this android can run faster on 4 limbs than 2? Humanoid forms just aren't designed to work this way, so unless it's somehow able to transform into something resembling a horse? Even then, some impressively wierd engineering would be needed; 4-legged animals tend to have "reverse jointed" rear legs which are the same length as their forelegs, so unless this is some sort of T-1000 "liquid metal", it'd have to both rotate its rear kneecaps and stretch it's arms...
Unless of course, it doesn't look human. But then, it's not a "she".
Which takes us to the next fun bit: how would a modern military deal with a super-tough, super-smart android? Assuming of course, that it actually decides to get involved with physical combat.
And for this, I'll lean once again on the Terminator movies a bit. And I'm going to assume that the android is going to want to physically fight, despite the fact that it's meant to be super-intelligent.
Against small groups of human soldiers, then the android is probably going to able to run riot. And if the android has any sense, it'd go for divide-and-conquer style tactics to whittle the enemy down, or target their senior command. Maybe even raid their capital city to kill or threaten the political leaders (though see previous comments about how they're likely to have their own android guards...).
But even then, once the military realises it exists, it wouldn't be that hard to set up a killzone which negates it's advantages.
After all, it can't be everywhere at once; any sufficiently "evil" military would simply send in a hard massed assault on the village. Because then, you either draw the android out and destroy it in direct combat, or you force it into a situation (e.g. by taking hostages) which funnel it into a tight space which negate it's capabilities.
You could even just send in a tank or some bombers or even a cruise missile to eliminate the village, unless they have significant radar and missile defence technology and the infrastructure to maintain them. And at that point, are they really a village?
And what happens if/when the village is captured? Either the scientist is forced to order it to surrender, or the army can simply destroy the village; what happens to the android's "super-intelligent" motivations at that point?
Beyond that, a humanoid robot with an endoskeleton isn't going too be too hard to take down with modern military weaponry: you can watch something like the TAOFLEDERMAUS channel on Youtube to see how even "home-made" 12-gauge bullets can chew chunks out of plate steel, never mind military armor-piercing shells and the like. And at 160lb, this android sounds too light to have any significant amount of armoring.
There's also the question as to how much damage an android could actually take. Obviously, it could handle the loss of one or more limbs far better than a living creature could, and any designer worth their salt would put the CPU and power source in the chest, and shield them as much as possible. But this also means that any body shots are highly likely to take the android down, especially one as lightly armored as this one. And that's before you consider what happens if any support systems are damaged, such as cooling for the power supply. Or what happens if someone gets a head-shot in? One weakness of the humanoid form is that most of our sensors are in the head; how many useful sensors can you install elsewhere in a "human" android which wouldn't make it look inhuman?
Conversely, you've got the T-1000 liquid metal/nanotech approach, and a machine which could literally dissolve into the ground, seep through any cracks and effectively ignore bullets (to a point). But at that point, we're getting even deeper into science-fantasy.
Fundamentally, and no matter how powerful, any individual trying to defend a static position against a large and equally capable army is going to lose. Because they can't be everywhere at once, and they're constrained in what they can do.
Unless you do a lot of handwaving, as per most fiction involving military robots ;)
[*] I know Roy dips his hand into liquid nitrogen in Bladerunner, but this is a fairly big bit of handwaving, given that the entire premise of Bladerunner is that androids are physically indistinguisable from humans and can only be detected by psychological tests. A simple DNA test or chemical scan should have been able to detect physical capabilities such as these...