Interglacial Crash

In the last Interglacial the sea level rose steadily, and other indicators confirm that it kept getting warmer. Then at the end there was a relatively quick change. Sea level started falling at 1cm per year: which is a lot of ice building up somewhere. I've never heard of a climate simulation that demonstrated this crash. Well here's a picture of the current heat anomaly that I've nicked (the link to) from John Baez's Diary:

If the sea gets warm (and big) leading to lots of moisture [notwithstanding the silly predictions of warming leading to drought] then you get a lot of snow. Then maybe at some point the CO2 negative feedback starts to kick in [and they must exist: to claim otherwise is to claim that the previously stable CO2 level was an unstable equilibrium], CO2 drops but the water remains warm. Then maybe the sun goes quiet or there are a few big volcanoes in a row ["year without a summer"]. The snow persists and reflects. All seems too easy...