- Compact style
- Indico style
- Indico style - inline minutes
- Indico style - numbered
- Indico style - numbered + minutes
- Indico Weeks View
This periodic open meeting of the LHC TOP Working Group takes place twice a year at CERN. The meeting provides a forum for discussion between experimentalists and theorists on the latest developments in top quark physics. Everybody is welcome to attend; please register using the link on the left.
Vidyo will be available
ATLAS ttbb:
- Fusing algorithm also applicable to many other processes, including V+jets, dijet + jets. Yes, this was first developed for Z+jets, see slide 20.
CMS 4-top:
- p3: difference between two theory calculations is mainly driven by scale choice, not by inclusion of EW NLO. ATLAS is also happy with this second calculation, we should agree on theory prediction to use within LHCtopWG.
- Scale choice for experimental MC is the same between ATLAS and CMS.
- p11: Correcting Njets and Nb: Using two different CMS ttbar predictions, different one for 2016 dataset and 2017+2018 dataset. Njets distribution was perfect for previous dataset, but not for current one, that’s why there are corrections here.
- bb correction is largest uncertainty in the end (see page 14)
- main generator for tt, ttbb, ttW/Z/H used for this analysis aMC@NLO with FxFx. Negative weights not an issue since this is not in the high-mass tails. (for ttbar analyses, default generator is still PW+PY).
LHCb update:
- What physics can be done with large 300fb-1 dataset? Exploring other top measurements, more than just XS and asymmetry.
- Define common fiducial region for LHCb/ATLAS/CMS? Yes, but would have to be based on leptons, or be relatively loose.
ATLAS spin correlation:
- Green line is weak corrections, but several other changes in this calculation compared to NLO curve, see slides 27, 28
CMS spin correlation:
- difference between ATLAS and CMS? data look similar, PW looks similar, NLO+weak corrections agrees better with data for both experiments.
- CMS doesn’t have MCFM comparison (ie. NLO in decay)
- ATLAS doesn’t have aMC@NLO with FxFx
- Since this is a ratio, expansion requires careful consideration of LO and NLO effects in numerator and denominator, see slide 23.
- In principle, D is the most sensitive variable, but this requires boosting into the ttbar rest frame, thus less sensitive than Delta Phi in the end.
- CMS finds same f_SM value as ATLAS with Zongguo’s NLO+Weak calculation.
Spin correlations at NLO+weak corrections:
- weak corrections, not electroweak
- Difference between their prediction and Powheg? PS and Hadronization, but these effects should be small since top is not reconstructed.
- Would be very useful to document this calculation and the settings used. So far citations used by experiments are several years old.
- Running with a dynamic scale is difficult for the authors due to numerical issues in the calculation.
Spin correlations at NNLO:
- There is another calculation of EW corrections, see p4, could ask these authors for their calculation
- Would be nice to have Bernreuther/Si calculation in fiducial phase space to compare on p1
- Can implement proper ratio calculation as shown by Christian on p23? Yes, can work on this.
- Asymmetry in the NNLO prediction? Result of dynamic scale.
- Using fixed scale of mt/2 is preferable over mt, and calculation has shown that this should get even closer to the data. Dynamic scale relevant for high pT, high mass, but this region here is dominated by bulk.
- Update after the meeting: Expanding the ratio makes a big difference when going from NLO to NNLO: http://www.precision.hep.phy.cam.ac.uk/results/ttbar-decay/
Common MC:
- p14: no requirement on b-jet matching, but any such requirements makes the difference with/without Evtgen less.
- Impact of Evtgen: this should not produce any difference, radiation for b quarks (mass of 5 GeV) should not even be at the level of a few %. Is it possible that there is a coding effect or something specific to the implementation? Could it be due to neutrinos? - Can check
EFT 4-fermion interactions:
- this is recasting an ATLAS top to Zq search to look for 4-fermion interactions, thus looking at the control region of the ATLAS analysis
- p27: how did you do this recasting? EFT to actual Z’? - start from Z’, integrate to find operators that contribute, then obtain limits
- Kinematics of lepton from 4-f top decay also different? Yes, but mainly due to invariant mass effect, not helicity of top
EFT recent theory progress:
- p3: reweighting would also require input events that were cut by the original analysis. Interesting proposal, but it can be complicated depending on how the experimental analysis is done (profile likelihood fit or complicated MVA that relies on complicated correlations).
- p17: sensitivity of VBF process? See VBF column, this would be more promising than eg ttVV.
Top mass theory:
- p10: R is the radius of the jet
- p17: first bullet should say “In a fixed-order” rather than “In an NLO..."
Not clear how to do the correction from long-distance to short-distance mass for unfolded distributions.
- linear power corrections depend on method used to measure the mass - this could reveal size of the corrections, or show that these are small.
- p18: What happens when one goes away from boosted regime? Not clear.
CMS top mass:
- measurement itself sensitive to input top mass? No, see page 15
- check if excluding first bin where threshold effects are important? Would also lose sensitivity here. Also, trends seen on p15 are due to mass but also alpha_s and PDF, see p19 when this is fixed.
- p19: What type of fit is it? Chi2 minimization with all uncertainties included, about 20 free parameters, with more than 1000 data points (20 for ttbar, 1000 for Hera data).
ATLAS top mass:
- threshold behavior in MSbar running mass scheme - why such a large uncertainty? Due to larger change in observable in this case - sensitive near threshold
- Comparison to off-shell calculation - done at one mass point due to limitations of computing resources, then showed good stability as function of number of bins (p20)
- At 13 TeV, ATLAS and CMS analyzers are in contact with each other and theorists with regards to the theory calculation
- Statement that this is the most precise pole mass measurement - several different measurements available now, probably makes sense to be more precise in what type of measurement we are referring to (and this statement is not in the paper)
- That there are no higher-order corrections is likely true here because you are close to the threshold, close to the pole region. Could try different methods to address this question, scale variations alone might not be enough.
ATLAS ttgamma:
- p8 PPT is trained without isolation cut, and there is some residual correlation.
- p13: electron fakes scale factor - slightly above 1 in central region, larger values in forward regions
- Sensitivity to gluon-gluon vs quark-quark initial states - have not studied in detail, but enlarged in q-q right now. Hoping to enrich quark-quark through selection cuts, then eg look at charge asymmetry. If gluon-gluon initial state, then photon mainly coupling to top quark.
- Updated theory calculation for normalization in the next iteration will have dynamic scale choice that reduces scale uncertainty.
CMS EFT:
- p4: inclusive (or fiducial) cross-section measurements more and more rely on profile likelihood fits - very sensitive to difference in kinematics, even in fiducial region
- p4: EFT in direct measurements through reweighting - as long as fully parameterized in each bin, as well as all events that fail cuts.
- p15: This is a different reweighting procedure based on the distribution (what is called hybrid on p4)
- We generally should also consider limits obtained from b-physics. These constrain some top operators but require additional assumptions.
- in ttZ, can also constrain 4-fermion operators for which there is no direct limit yet.
Top width:
- note that resonance-aware matching in madgraph is not the same as this allrad feature, madgraph WbWb does not include radiation off a b from top decay at ME level (see p11).
- p16: normalized differential distribution, does the dashed curve cross 1 anywhere?
- Impact of higher order corrections, why smaller uncertainties with MG5? Seems the higher order corrections reduce the shift in the high tail, see summary table of systematics on p24. Radiative corrections on interference depend on width.
- Modification of total XS and width? Not used as a constraint, distributions are all normalized to this fiducial region
- Re-doing this analysis with full dataset has promise for significantly reduced uncertainties.
- p4: indirect extraction from single top assumes BR(top to Wq)=1, what is the impact of this assumption, can it be measured?
- This here is really BW shape, for width measurement of Higgs, interference enhances the sensitivity of off-shell region
Single top combination and Vtb:
- theory prediction used? See page 7 and 8, want to have consistent treatment of PDFs.
HL/HE-LHC:
- p13: improvement by order of magnitude going from today to 300 fb-1, another order of magnitude from 300 to 3000 fb-1
- p15: to get idea of signal statistics, multiply by bin width.
- p8/9: sensitive to q qbar production through high rapidity and high mass, but even at high m(ttbar), the antiquark is mainly at low x.
- Physics to be done with forward tops: Not just PDF (see p9), but also eg asymmetry and other measurements.