Faults not truncating

[s-pearsongrant]

We have recently found that only some faults truncate against younger ones. We have checked with your example 5 and similarly the first fault truncates, but when we add a second fault they cut through each other and both keep going. Below is the fault relation matrix and the resulting geological model.

This seems to be a fairly new behaviour in the last couple of months. All the faults are different series. Do you have any thoughts on how to fix it?

Many thanks,
Sophie

[s-pearsongrant]

For anyone else who might come across this issue, it appears to be related to PyTORCH. The most recent version (2.11) shows this behaviour, but reverting to version 2.8 of PyTORCH the faults truncate as expected.