Emacs-jupyter code-blocks and unicode completion


I’m trying to set up unicode completion in julia for emacs-jupyter and have almost got it working. However, I have a really weird issue. The unicode suggestions show up as expected, but when I select them (by tapping enter), the actual characters (say, \alpha) are inserted instead of the symbol for alpha. It’s really driving me up the wall and I can’t figure out what’s happening.

I have also noticed that if I go M-x company-math-symbols-unicode and then select alpha (by pressing enter), it inserts the correct unicode symbol. But this way, I will have to M-x every time I need to insert a symbol.

Has someone had a similar issue? Thanks for any tips/suggestions/comments.