After running multiple experiments on the function f(x), it seems that the function is selective for concepts related to popularity or mainstream relevance. Words like "Hollywood", "oxygen", and "rhythm" are associated with higher values closer to 1, indicating a stronger association with the concept. On the other hand, words like "death", "north pole", and "Bitcoin" have lower values closer to 0, indicating a weaker association with the concept of popularity or relevance. It's important to note that the function may also be selective for other concepts as well, but based on the given words, this description seems to explain most of the function's behavior.

Please let me know if there's anything else I can help you with.