The village from the Asterix and Obelix comic books were completely autonomous from the Roman empire even though the rest of Gaus was conquered. Of course that specific village is fictional, but I'm wondering if it is based on anything real. Were there any such villages in the Roman empire that were actually autonomous? Did the Roman state have any control of villages at the edge of the empire at all?
It depends on how you mean. Were there secluded villages in hidden valleys, cut off from the outside world, continuing the old ways like a Shangri-La? In that sense, no. When we think about the extreme amount of interconnectedness in the modern world we often contrast it with an image of older times in which the friction of distance was such that, effectively, a village could be its own world. There is some truth to this (will get to that later) but fundamentally it is wrong--even a theoretically self sufficient village of largely immobile farmers would still be bound by great chains of interconnectedness from village to village that, in aggregate, could stretch enormous distances. And of course said theoretical village and immobile farmers did not exist, individual mobility throughout history was actually quite high. To give an extreme end of the scale, the historian Walter Schiedel estimated that perhaps 40% of the population of early imperial Italy relocated some significant distance during their life, and that does not include temporary trips for commerce, official needs, religious travel, etc. As I say this is very far on the end of the scale, but at least illustrates that there is nothing inherently immobile in the ancient world.
Now that romantic interpretation out of the way, were there areas that, as the saying goes, the writ of the empire ran thin? Absolutely, yes, fewer than might be expected but there were definitely "lawless" areas. Perhaps the most notable example is Isauria in what is now southern Turkey, whose people had a reputation for fierceness and banditry, there is even a military campaign recorded in the late fourth century to "pacify" it. The fact that a campaign was required points to why the framing that ancient authors provides these latrones (bandits) is not really satisfying, highwaymen do not require military campaigns, this was clearly an area that became temporarily effectively independent--or more accurately, had always been effectively independent but because of changing circumstances the way that status was viewed from an administrative standpoint changed. It went from being a "wild area" to A Problem.
There were other areas like this, such as parts of Wales, the eastern desert of Egypt, northern Spain, places where the friction of distance becomes such the mechanisms of imperial administration simply cannot function and thus other forms of organization developed. And this has a long legacy--the interior of Sardinia is still today called "Barbagia". This effective independence might be interpreted through the lens of crime--so that Cicero called Sardinians "little sheepskin wearing bandits" (matrucati latrunculi)--but it is likely that something more systemically ordered and significant was going on.
I am not aware if there has been a full study of the phenomenon of these internal frontier zones, but Shaw's "Bandit Highlands and Lowland Peace" is a good study of Isauria specifically and its framework is widely applicable elsewhere. For a theoretical perspective however, James C Scott's The Art of Not Being Governed provides a powerful reason to see these sorts of areas as being zones of resistance.