when I sleep = when I don't tweet
Some automated app gimic scans your tweets and then tells you when you typically sleep. It also assumes that when you sleep you don't tweet. I assume that's a fair assumption in 2010, but I doubt such for 2015. Anyway, of course it also presupposes that when you are awake you are tweeting. Quite a premise, no doubt self-serving.
What's frightening here is not that they assume wrongly that I go to bed at 11 pm, but that I am up at 5. I am, typically, but does that mean I'm using Twitter that early? There's a thought. Overall my response is: no thanks.
A note on grammar: "more likely to sleep..." More likely than what or whom? More likely than otherwise?