Loose PID_MU cut removes nearly all signal events

In my HLT2 line I am cutting on F.PID_MU>0. Since Thursday 10.3.2022 this cut removes nearly all signal events when checking the efficiency of arbitrary Muon modes with MooreAnalysis. As an example: The efficiency on B0->Kstmumu simulation droped by 85% compared to the last evaluation before this bug occured on monday. When effectively removing the PID cut by setting the cut value to -10000 the efficiency looks just fine. Even a very loose cut of F.PID_MU>-10 reduces the signal efficiency by 75%, which leads me to believe that this is a bug. Reverting my stack to the status of Monday reproduces the results I saw back then.

My trigger line: !1224 (merged) lhcb-datapkg/ParamFiles!36 (merged)

MooreAnalysis output with broken pid cuts: out_moore_broken_muonpid.txt out_eff_broken_muonpid.txt

MooreAnalysis output with up to date stack but without muon pid cuts: out_moore_nopid.txt out_eff_nopid.txt

MooreAnalysis output with stack reverted to my last commit before the bug: out_moore_monday.txt out_eff_monday.txt