[BUG] Converting the Behroozi mass function to different mass definitions does not seem to work

[puchwein]

Describe the bug
Converting the Behroozi mass function to different mass definitions does not seem to work properly.

To Reproduce
Steps to reproduce the behavior:

hmf_test.py

1. Calculate Behroozi mass function using the SOcritical definition

2. Find corresponding overdensity value for the SOmean definition for the same actual threshold density

3. Calculate Behroozi mass function using the SOmean definition with the overdensity found above

4. The mass functions calculated in 1. and 3. differ by up to 50%.

Expected behavior
The mass functions should be identical, as is the case when, e.g., using Tinker08 instead of Behroozi.

• OS: MacOS 26.2
• Python 3.14.2
• hmf version 3.5.2

Additional context
Sample script for reproducing the issue is attached.

[steven-murray]

@puchwein I believe this was fixed in #281 but I'd really appreciate if you could check that your use-case works as intended!