Attendees: Ashot, Chao G, Chao P, Dipangkar, Haiyan, Maxime, Nilanga, Weizhi, Xinzhan, Li

Xinzhan:
- cosmic removed gem efficiency with more statistics explains cross-section drop at high angle
- still too much data loss (>90% after 4 degree)
TASK: improve neural network algorithm for less data rejection
- question about too high number of modules by cluster (19 to 29 for crystal)
TASK: investigate number of module versus module energy threshold
- phi dependency of gem efficiency show some discrepancies: can be due to trigger efficiency or/and different theta range of cyrstal/LG for different phi sectors.
-> low gem efficiency where high hycal yields -> need to clean up hycal

Maxime:
- problem with wiggling of 2GeV simulation solve with good calibration
TASK: points just after 2 degree for 2GeV show some data/simulation discrepancy -> need to study this
- GE preliminary fit between 0.7 and 0.9 fm for both 1GeV and 2GeV
TASK: extract hycal acceptance to be able to correct both simulation and data directly
- updated trigger efficiency with new island code match old primex one