Shu's weekly group discussion

Wednesday 25 Mar 2026, 10:00 → 13:35 Asia/Shanghai

Tsung-Dao Lee Institute/N6F-N630 - Smart Classroom (Tsung-Dao Lee Institute)

Matthias Tartarin (TDLI/SJTU)

点击链接入会，或添加至会议列表：

https://cern.zoom.us/j/4042031778

Introduction & Announcements

Meeting's note: From Prof. Li's general advices during the meeting (paraphrased by M.Tartarin):

• When writing your minutes, if you present multiple analysis, please put them in decreasing order of personal contribution 
  (i.e. 1] Contribution +++ ; 
         2] Contribution ++ ; 
         3] Contribution + ;
              ... ).
• If you present a (new) work (or if it is rather technical), make sure you present the scope and the obectives to the team members. Either with a short presentation in a few slides, or inside the minutes. 
   

---

Today's focus: 

[1st project/analysis: XXX ]:

• item 1
  • sub-item 1
  • sub-item 2
• item 2
• [...]

[2nd project/analysis: XXX ]:

• item 1
• item 2
• [...]

[...]

---

Plans: 

• [optional: indicate which project/analysis it refers to] item 1
• [optional: indicate which project/analysis it refers to] item 2

---

1 Status

Speaker: Kang Yu (TDLI, SJTU)

2 Status

Speaker: Linjing Duan (TDLI, SJTU)

Today's focus:

[CEPC Neutral Triple Gauge Couplings (nTGC) in the Zννˉγ Channel]:

• Writing the simulation part of the paper draft, here is the summary of this part:
  • The study targets the EFT sensitivity to 9 nTGC operators at CEPC.
  • The signal cross section is parameterized as SM + interference + quadratic terms; in 2D, an additional operator-mixing cross term is included.
  • Signal samples are generated with MadGraph + Pythia8 and simulated with CEPCSW.
• Current task: redo the 2D fit for the EFT operator coefficients and optimize the scan range to obtain more sensible confidence contours.

[ZZy]:

• Studied the paper and the supporting note, and now have a clearer understanding of the physics goal and the main analysis strategy of this project.

Plans:

CEPC:

• Continue the 2D fit optimization for the EFT operator coefficients.
• Help Yuning investigate the "0 hit" issue in the e+e−→ννˉγ MC samples generated by MadGraph in the full simulation chain (sim, tracking, digi, and reco).

ZZy:

• Discuss the next concrete steps of the project with Matthias tomorrow

3 Status

Speaker: Jun Wei Lok (TDLI, SJTU)

Ansys heat simulation

•  finalizing the meshing stage for the ECAL thermal model to ensure high-quality cell distribution.
• will begin the solver iterations once the mesh optimization is complete.

4 Status

Speaker: Zelin Yan (TDLI, SJTU)

Today's focus: 

[H2a4b analysis]:

• Update new coffea code for SR2b3j and CR0b
• Update baseline cut for SR optimal and Z+jets HF CR selection

[HGTD monitoring ]:

• Link OPC-UA server with remote FELIX card
• Monitor vars on UaExpert and update with time

[Hyyd analysis] :

• New exp(obs) limit updated after fixing bug in  e->y  with pt > 200GeV:  0.92%(1.16%) -> 0.94%(1.10%)

---

Plans: 

• H2a4b analysis: 
  • Discuss with Bingxuan about Z+jets HF CR selection and prepare for new status report in HLRS meeting
• Hyyd analysis: 
  • Combine the result with Run2 using new bug fixed ws
• Wlllv analysis: 
  • Update new SR definition in BDT training 
• HGTD monitoring: 
  • Optimize monitoring design and configs 
  • Add  more vars into monitoring 

5 Status

Speaker: Bai-Hong Zhou (TDLI, SJTU)

Today's focus: 

[ssWW]:

• Give a presentation for the WZ CR on ssWW weekly meeting;

[Hyy*]:

• Derive the fit with each category separately;
• Update the correlation schemes (decorrelate muon related NPs, and correlate EG_resolution);

 

---

Plans: 

 

[ssWW]:

• Continue the harmolization study for production framework between easyjet and VBSTools;
• Make sure my MR for the WZ CR would be approved;
• Compare the renormalization and sample selections between Rel 21 (Sherpa 2.2.2) and Rel 22 to see if there are any differences;
• Check the selection criteria in the WZCR between Rel 21 and Rel 22;

[Hyy*]:

• Wait for the updates from the background function study;

[Zbbj Calibration]:

• Perfrom with the stat only fit cross check;
• Work with non-prompt fit;
• Work with Changqiao for derive the final scale factor;

6 Status

Speaker: Xuliang Zhu (TDLI, SJTU)

7 Status

Speaker: Zhiyu Zhao (TDLI, SJTU)

Today's focus: 

DarkSHINE:

• v1.6 analysis
  • Using the BDT method, I performed a preliminary study of the signal and background yields for samples produced with the new digitization algorithm. Compared with the previous truth-level analysis(left), the signal efficiency becomes slightly worse after introducing digitization(right).

• ECAL prototype
  • Finalized the basic external mechanical dimensions of the 3×3 crystal array, and 3D-printed a mold to test the tolerances; look fine.

PhD thesis:

• Finish the draft of the main body of my work (5/8 chapters, text only). I am currently doing some additional studies for supplementary content, starting with the impact of detector resolution on the signal-to-background ratio (S/B). In addition, I have also started reading some review articles on dark matter theory and experiments.

---

Plans: 

 

DarkSHINE:

• v1.6 analysis
  • Filtering out feature variables with high correlation and low ranks
  • Think about how to optimize difitization algorithm
• ECAL prototype
  • Assisted USTC people in completing the design of the front-end electronics board

8 Status

Speaker: Haidar Mas'ud Alfanda (TDLI, SJTU)

H4b analysis

• Process of interest: ZH -> ll h2(h1 h1 -> bbbb), light singlet scalar mass range ms = {15...60...5}
• Isolate leptons from Z decay
• 
• Reconstruct 4 jets exclusively using eekT algorithm
• The figure below shows the comparison of reconstructed invariant mass distributions from the current analysis (left) compared to the previous one (right) 
• 
• To identify the origin of jets, we used a ParT model trained on ZH -> nunu H -> qq/gg for jet tagging
• However, events with all 4 b-tagged jets is about ~53%
• Next plan:
  •  Fine tune the jet tagging model. If it proves not very helpful I will switch to the Foundation Model
  • I found a bug and fixed it in my code to investigate jet merging at lower scalar masses
  • Quick checks indicate that above ms = 25 GeV, jets are resolved
  • Produce ntuples for the background processes

9 Status

Speaker: Matthias Tartarin (TDLI, SJTU)

Today's focus: 

[ssWW Analysis]:

• mc23 samples (post easyjet-fix by Baihong): Done producing, last one was on Monday monring so I was on time.
• Baihong finished with mc20 earlier. We cross-checked each others (good thing, trouble results errors at first on my side with FastFrames).
• Very nice presentation from Baihong to the GM about these results. Some directions were suggested by the group to investigate WZ_QCD and WZ_EW : Will be important to some extend for the reweighting too.
• Speaking of which: Very short window, but crossed my fingers and ran the reweighting script built weeks ago on these new mc20 and mc23 samples: Great improvement on the fit, comparison now possible with legacy plots.
• Presented this part to the GM yesterday: Got good feedback from the team, in particular was very interesting to hear from those who contributed to the legacy analysis. I have (/will have) directions to explore to improve even more than what has been done for the legacy.

[Hyy*]:

• The plots I produced for the fakes were did things differently from what was expected. Was not very surprising, it was hard to fully understand from the legacy note. But comments from AC during these past days were extremely informative.
  • looking forward to the discussion next meeting tomorrow. 
•  Danning did a lot of work on background modelling, she will present tomorrow during group meeting. I parsed the slide she put on Mattermost days ago, and it looks like the issue Laura encountered is now solved (i.e. by narrowing the fitting range). So all good?

---

Plans: 

• [ssWW] Work on the latest opened questions raised yesterday during GM about the reweighing (and the yield too?). Time to also start looking at other tasks. Baihong and I have our eyes on "fakes" (already 2 people working on this) and "polarised samples using MG" (Yi worked on this a year ago, she will come back in April, check with her to see if this could be worth updating or exploring again) -> check status on these task ; start contributing where we can. 
• [Hyy*]: Continue working on the (correct) implementation of the fakes. Wait for the GM, see from there.
• [ZZy]: Assess the status on this analysis, during and after the meeting with Prof. Li and Linjing.

REU: Undergrad Research Discussions

10 Status

Speaker: Peixu Teng (SPA, SJTU)

• Weekly Summary
  • Heavy neutral leptons
    • Basically sorted out the issues related to LEP in MadGraph.
    • 
      • 
      • Here, i approximate Br to 1. P_dec is the decay probability of a heavy neutrino within 1 meter. L_int​∼200 pb^−1
      • cut:∣cosθγ​∣≤0.7,Eγ​≥0.7 GeV.
    • I found that this can be verified using an analytical solution method, which I am currently working on.
  • Detector Simulation
    • On my to do list

11 Status

Speaker: Zhijing Zhang (CCNU & TDLI/SJTU)

eg.energy resolution of 1 GeV.png

Today’s focus:

• Learn how to perform optical simulation of a single crystal.
• Use 0.1 GeV as a step size, simulate the energy, position, and time distributions of gamma from 0.1 GeV to 3.5 GeV at a fixed theta angle of 90 degrees. The mean and sigma for each energy point were obtained and finally compiled into a single figure, showing the curves of energy, position, and time resolution as functions of energy.

 

---

Plans:

• Repeat the above steps to perform full-space angular simulation,  analyze the resulting plots, and compare them with the fixed-point simulation.
• Conduct beam test simulation by modifying the particle gun settings, then compare the results with experimental data(preliminary plan).

12 Status

Speakers: Jiacheng Shi (SPA, SJTU), Yuning Chen (Zhiyuan College, SJTU), Zhengyang Qian (SPA, SJTU)