Bugfix for a SUSY JO control file; higgsino neutralino pair production decaying into bino LSPs

The sign of one mass parameter was incorrectly set in the current control file (SUSY_SimplifiedModel_N2N3.py), which biases the kinematics of generated SUSY particle much harder than they should be (see page 6-7 of this).
The sign of neutralino3 mass is fixed to be negative to be consistent with the mixing matrix configuration intended initially.

However then the there turned out to be the second bug that the mixing matrix information was not provided to the job at the first place, and the default wino-NLSP/bino-LSP mixing had been just used. (This is an unfortunate incident caused in the Rel 20.7->Rel 21.6 transition where the mixing parameters were defined in the separate param_card files in Rel 20.7 while this has to be explicitly written in the control files in Rel 21.6)

Some quick validation has been done after those fixes:
HiggsinoBinoNNGridReq_20201103.pdf
The sign fix corrects the kinematics toward the expected direction.
The mixing matrix change actually doesn't affect anything on the kinematics (the change is confirmed in the log).

The CIs have been skipped as they don't seem to be very relevant for this kind of fix (but please correct me if this is wrong)