Fix bug when reusing jax logp for initial point generation

[ricardoV94]

Fixes issue described in https://discourse.pymc.io/t/attributeerror-numpy-ndarray-object-has-no-attribute-at-when-sampling-lkj-cholesky-covariance-priors-for-multivariate-normal-models-example-with-numpyro-or-blackjax/16598/3

Caused by #7681
CC @nataziel

---

📚 Documentation preview 📚: https://pymc--7695.org.readthedocs.build/en/7695/

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.64%. Comparing base (53e0774) to head (9ad9cf7).
Report is 63 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #7695      +/-   ##
==========================================
+ Coverage   90.72%   92.64%   +1.92%     
==========================================
  Files         107      107              
  Lines       18324    18324              
==========================================
+ Hits        16625    16977     +352     
+ Misses       1699     1347     -352     

Files with missing lines	Coverage Δ
pymc/sampling/jax.py	94.11% <100.00%> (-0.91%)	⬇️

... and 13 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[nataziel]

Just for anyone investigating this later I think this was directly caused by #7610, not #7681