Every claim this atlas records about Charvaka arrives secondhand, filtered through opponents who preserved the school's arguments only well enough to knock them down, which means reconstructing the position at all is a discipline of its own, closer to detective work than ordinary textual study. Scholars read a Vedantin's paraphrase against a Jain's paraphrase of the same lost verse, weigh which details recur independently in hostile sources that could not have copied each other, and treat agreement across enemies as the nearest thing to evidence a materialist voice was ever really making the argument attributed to it.
The clearest test case is Jayarashi Bhatta's ninth century Tattvopaplavasimha, The Lion That Devours All Categories, which does not simply assert that inference is unreliable but works through each accepted means of knowledge in turn, perception, inference, testimony, comparison, and shows that every proposed defense of it depends on the very faculty being defended, a structure of argument later readers have compared to Hume's observation that no finite run of past cases can license a claim about cases not yet observed. Whether Jayarashi meant this as materialism's own weapon or turned it against materialism as well, denying that even Charvaka's atoms and elements survive scrutiny, is still argued over, and the disagreement is instructive on its own terms: it shows a text can be read as either the last word of a school or the proof that no school's foundations hold, and the surviving evidence does not settle which.