The other half is how you find them, because searching by keyword produces exactly your folder - a large pile with no structure and no sense of which items matter.
Follow citations instead. Find two or three papers that are clearly central to your topic, then work in both directions:
Backwards, through what they cite. This takes you to the foundational work and shows you the lineage.
Forwards, through what cites them. This takes you to the current conversation, and it is the direction people neglect. Anything highly cited that came after your central paper is something you need to know about.
Do that for a while and you will notice the same names and the same handful of papers appearing repeatedly. That recurrence is the signal. Those are your pass three papers, and they identify themselves rather than requiring you to judge.
A good marker of progress: when you can look at a new paper's reference list and recognise most of it, you have the shape of the field. That usually arrives at a few dozen papers read properly, not two hundred.
Also worth finding early: a recent review article in your area. Somebody has often already done the mapping, and a good review saves months. It is not a substitute for reading the primary work, and it is an excellent index to it.