This is my first time using hierarchical schematics but I believe this is a bug.
If I delete the LT1011 comparator in the schematic then everything runs fine. With the comparator inserted, the simulation fails.
I had a larger schematic that was working and decided to clean it up by using hierarchical schematics (for my first time).
When it did not work, I started deleting components until I had the smallest test case possible.
The strange thing is - I had a LT1195 opamp in that larger schematic as well as the LT1011 and that LT1195 does not affect the results at all.
Thus I suspect some strange interaction that is unique with only the LT1011 model.
Attached are screenshots and files.
Here is the output with the LT1011 in the schematic. The inverter output is wrong and does not change.

And here is the output with the LT1011 removed. The inverter output is as expected.

Here are all three files used
All 3 files are in the zip. In case it makes a difference, the top level file "Draft8.asc" was in a different folder than the inverter symbol and schematic.
Everything runs without errors and as you can see I can open the files properly by right-clicking the symbol in the top level schematic.
I tried renaming the V+ and V- pins and net labels to vcc/vee hoping it was the names I used - but that had the same result.
Any insights would be appreciated.