This is another one of those small signs that the combination of ice thickness and warm waters are overruling the influence of weather conditions. For the first time this melting season the IJIS numeric data of sea ice extent file is showing an uptick for yesterday's extent number after revision (albeit a very small one of just 938 square km).
To show how late this is, I've made a print screen of my spreadsheet:
Here we see daily extent decreases for a particular date in the 2005-2011 period. The green cells are for first days showing an increase in extent at the end of the melting season (yellow is for century breaks).
I can't really say the weather is to blame, as there hasn't been a real Beaufort Gyre or Dipole Anomaly set-up for weeks. It must be the ice then.