Difficulty in Finding gg->ZZ* continuum in Monte-Carlo

guest_hkeoshkerian · 2 October 2026 05:59

Hello,

I’ve been trying to locate the gg2ZZ samples in Monte-Carlo in the 2025 ATLAS Metadata. Where can I find these samples? I have been running an H->ZZ->4l analysis, and have an overall discrepancy of ~1.4 outside the 115-130 GeV mass window.

For the MC samples, I’m using the ones listed below, but I think I’m missing some samples.

Signal

DID	Catalogue name
345060	`PowhegPythia8EvtGen_NNLOPS_nnlo_30_ggH125_ZZ4l`
346228	`PowhegPy8EG_NNPDF30_AZNLOCTEQ6L1_VBFH125_ZZ4lep_notau`
345066	`PowhegPythia8EvtGen_NNPDF3_AZNLO_ggZH125_ZZ4lepZinc`
346645	`PowhegPythia8EvtGen_NNPDF30_AZNLO_ZH125J_Zincl_MINLO_shw`
346646	`PowhegPythia8EvtGen_NNPDF30_AZNLO_WpH125J_Wincl_MINLO_shw`
346647	`PowhegPythia8EvtGen_NNPDF30_AZNLO_WmH125J_Wincl_MINLO_shw`
346340	`PowhegPy8EG_A14NNPDF23_NNPDF30ME_ttH125_ZZ4l_allhad`
346341	`PowhegPy8EG_A14NNPDF23_NNPDF30ME_ttH125_ZZ4l_semilep`
346342	`PowhegPy8EG_A14NNPDF23_NNPDF30ME_ttH125_ZZ4l_dilep`
346414	`aMcAtNloPythia8EvtGen_tHjb125_4fl_ZZ4l`
346511	`aMcAtNloPythia8EvtGen_tWH125_ZZ4l`

Irreducible background

DID	Catalogue name
700600	`Sh_2212_llll`
700587	`Sh_2212_lllljj`
700591	`Sh_2212_lllljj_Int`

Reducible and rare background

DID	Catalogue name
410470	`PhPy8EG_A14_ttbar_hdamp258p75_nonallhad`
410155	`aMcAtNloPythia8EvtGen_MEN30NLO_A14N23LO_ttW`
410218	`aMcAtNloPythia8EvtGen_MEN30NLO_A14N23LO_ttee`
410219	`aMcAtNloPythia8EvtGen_MEN30NLO_A14N23LO_ttmumu`
410220	`aMcAtNloPythia8EvtGen_MEN30NLO_A14N23LO_tttautau`
410408	`aMcAtNloPythia8EvtGen_tWZ_Ztoll_minDR1`
412043	`aMcAtNloPythia8EvtGen_A14NNPDF31_SM4topsNLO`
700601	`Sh_2212_lllv`
364242	`Sherpa_222_NNPDF30NNLO_WWW_3l3v_EW6`
364243	`Sherpa_222_NNPDF30NNLO_WWZ_4l2v_EW6`
364245	`Sherpa_222_NNPDF30NNLO_WZZ_5l1v_EW6`
364246	`Sherpa_222_NNPDF30NNLO_WZZ_3l3v_EW6`
364247	`Sherpa_222_NNPDF30NNLO_ZZZ_6l0v_EW6`
364248	`Sherpa_222_NNPDF30NNLO_ZZZ_4l2v_EW6`
700320	`Sh_2211_Zee_maxHTpTV2_BFilter`
700321	`Sh_2211_Zee_maxHTpTV2_CFilterBVeto`
700322	`Sh_2211_Zee_maxHTpTV2_CVetoBVeto`
700323	`Sh_2211_Zmumu_maxHTpTV2_BFilter`
700324	`Sh_2211_Zmumu_maxHTpTV2_CFilterBVeto`
700325	`Sh_2211_Zmumu_maxHTpTV2_CVetoBVeto`

I appreciate your help in this matter.

Thank you.

Best,
Houry

zmarshal · 5 October 2026 21:50

Hi @guest_hkeoshkerian ,

Thank you for the note! Are you following our standard H>4l workflow from here?

We don’t see such a large discrepancy outside the Higgs mass window in that analysis, so it could be something like fake leptons or trigger acceptance that is throwing you off. We don’t have gg2VV samples as a part of the standard Open Data release; gg2VV isn’t recommended within ATLAS any more either (we’ve moved to gg4l for Run 3, but in Run 2 that wasn’t recommended either).

Cheers,
Zach

guest_hkeoshkerian · 6 October 2026 05:26

Dear Zach,

Thank you very much for your reply.

We did indeed follow the instructions in the notebooks. The only issue we encountered is that, at one point in the code, the absolute value of the Monte Carlo event weights is taken, as shown in the picture below.

My understanding is that Monte Carlo weights should generally be used with their signs preserved. When we removed the absolute value and retained the signed weights, discrepancies appeared in the resulting distributions and yields. I initially thought this might be due to a missing Monte Carlo sample.

However, if the gg2VV sample is not available and its use is also not recommended, then it seems that the discrepancy may instead be related to the use of the absolute value itself. I would like to understand the reason behind the calculation shown below, taking the absolute value of the MC weights in the numerator, and whether this is intentional for these Open Data samples.

Best regards,
Houry