Review of FTR-18-018

Comments from CCLE, Kevin Pedro

Comments on V12

* full pythia 8 version for sample 3?

Added the full release version 8.212

* multiple definitions of HT - propose H_{T}^{trk} in sec 3.1 I changed the text describing the vertex finding to just use sum Track pT, then HT is defined in the trigger section

* trigger threshold vs. jet threshold not clear in sec 3.3 I clarified this so that threshold and jet pT threshold are clear for the two triggers.

*figs 4,5: z-axis has no label Added varepsilon to denote efficiency

figs 6,8: m_{H} = 125, m_{$phi} = 30 GeV (subscripts, spaces, comma, units) ctau = 0 cm (unit) Edited to reflect these changes

fig 8: (14 TeV) in H(125) plots slightly too far to the right, should be aligned with plot border like other plots Edited to reflect these changes

fig 9: H(125) -> (space before arrow, also for H(250)) Edited to reflect these changes

use B instead of Br Edited to reflect these changes

sigma_{H(250)} B (remove \times) Edited to reflect these changes

m_{#phi} = 15 GeV (space after equals and space before unit, all instances) Edited to reflect these changes

* off-pointing: not pointing back to the PV? might merit a bit more explanation I added a few lines to explain off-pointing

* full pythia 8 version for sample 3?

Yes that is correct

* "Track finding at L1 relies on selection at the front end" -> not clear what "at the front end" means

I added front-end electronics, then the selection is defined in the next few sentences about stubs

* might be useful to have a diagram/visual example of the algorithm in section 3.1

Since Track jets are used but not the main topic of the yellow report we don't want to emphasize it in a figure

* multiple definitions of HT - propose H_{T}^{trk} in sec 3.1

We define it once as sum track pT. Then in the text what is described is the track selection and jet selection to reject fake tracks that enter the trk pT sum.

* trigger threshold vs. jet threshold not clear in sec 3.3

The thresholds depend on which trigger we consider. Two L1 trigger thresholds are shown in Figure 2 one based on jet threshold and the other based on HT threshold. These are defined in parentheses in Section 3.3

* \ET vs. \pt used inconsistently in some places - is it always \pt?

Yes we changed the ET mentioned in the intro to \pt * Sec. 4.4 could probably be combined with Sec. 4.3 Done

* figures need to be provided as pdf Done

* fig 1:

Put "Jet Finding Efficiency" on y-axis, remove from pave text on plot Done

header should say "Phase-2 Simulation Preliminary" Done

x-axis should indicate that pT is gen-level

legend: remove space in "Fast Jet" Done remove "\Delta"

"soft. emul." is never defined - change to "emulation" and add some note in the text about evaluating the algorithm via software emulation Done Removed

* fig 2: header should say "Phase-2 Simulation Preliminary" legend: "Fast Jets" -> "FastJet" Done

use () or [] consistently for units on x,y axes Done

* fig 4, 5: header should say "Phase-2 Simulation Preliminary" y-axis and z-axis labels should have same font size use () or [] consistently for units on x,y axes Done

* fig 6: header should say "Phase-2 Simulation Preliminary" remove extra spaces in "(14 TeV)" use () or [] consistently for units on x,y axes y-axis title should be right-aligned legend: H -> #phi #phi: m_{H} = 125 GeV, m_{#phi} = 15 GeV (etc.) (consistent capitalization, spacing, subscripts, units) ctau = 0 cm ctau = 1 cm ctau = 5 cm (consistent spacing, units)

* fig 7: header should say "Phase-2 Simulation Preliminary" y-axis title should be right-aligned remove legend border/shadow add some color to lines use () or [] consistently for units on x,y axes Done

* fig 9: header should say "Phase-2 Simulation Preliminary" y axis title: Events / 3 ab-1 (add space) use () or [] consistently for units on x,y axes legend: use italic B in ROOT TLatex (in lieu of mathcal{B}) m_{#phi} = 30 GeV (etc.) (subscript, spacing) DispHT -> Displaced H_{T} HT -> H_{T} align L1 Rate... with rest of legend

Done

ARC comments (PAS v11)

comments from Andrei Gritsan

- I have a feeling that plots are not embedded in vector graphics, like .pdf file. If this is the case, is it possible to switch to .pdf files? Moreover, several of my comments on the figures are not addressed, such as removing jargon (what is TTBar? on Fig.1), unify the style (Fig.2 is very different), etc.

we made an effort to make the style more uniform and got rid of the jargon. Most figures are indeed .png.

- Write “Figure" when it is the first word of a seance, "Fig." otherwise. There are too many examples where this is violated, just to a search for "Fig." through the whole text.

Done

- You do not need to define (HL) and (BSM) in the abstract, leave this to introduction. Moreover, you are not consistent: SM is not defined in the abstract (ok), but HL is not defined in the intro (not ok).

Done

- Title: why do you limit yourself to the Exotic Higgs Decays in the title? Your work has a much wider scope. I understand this is one example, but you say it yourself on line 25 that application is much wider. I think it is to your benefit to advertise it wider.

Yes, it has wider applications and we changed the title accordingly.

- line 31: here or earlier, why not add a ref. to the CMS Phase II detector TDR or something like that?

- line 115: add space before units.

Done

- lines 135-136: Br and sigmas are not defined. However, in the second part all you are saying is that cross section of pp->H(250)->phiphi>4j is set to be the same as pp->H(125)->phiphi>4j. Just say so. It takes awhile to figure out what you mean.

We changed the sentence it, please let us know if anything needs to be changed.

- Conclusion: This is a new part, completely redone. However, I think it is not to your advantage to focus on some technical details of the outcome of the study. These are good illustrations which you should have, but conclusions, especially itemized with bullets should be about the bigger picture. Something like this (not as an example, but about the information to be passed):

=> without such L1 trigger, we would miss on many interesting physics channels, for example such and such.

=> with this trigger we can get such and such things, including >10 events of such and such, which would be enough for the discovery if Br is larger than this.

=> the extended trigger option will make it even much much better, with such and such examples.

We tried to rework this a bit (also, AMM did not like bullets, so we got rid of those)

Comments From Anne-Marie Magnan

Title: "Feasibility studies of a first level track...." --> or equivalent.

Abstract: "The CMS detector....track trigger. A potential increase in the CMS sensitivity....."

Last line: heavy Heavy -> heavy

Done

- everywhere relevant Heavy -> heavy (l.113, 130, 136)

Done

l.20 jet lifetime tag -- Add "for tracks with pseudorapidity |eta| < 1.0" ---> otherwise the future work about extending to outer tracker cannot be understood.

Done

Before section 2 / line 31 : add a short paragraph about the CMS Phase 2 detector, with references to all TDRs, bibtex pasted below [*] for your convenience.

Done.

l.36 impact parameter in the x-y plane (d0)--> add "in the x-y plane"

Done

l.38-40 please add exact Pythia v8.205 and Powheg v2.0 versions.

Done

l.67 add "ones" --> much less pure than low pT ones

Done

l74 as noted previously, please introduce properly HT and quadjet triggers !! Add a full sentence beforehand for each, with definitions of trigger variables and selection thresholds involved.

Done

Captions fig1 and 2: last sentences for each, it is unusual to give comments about results in a caption, these should be moved in the main text.

Yes, We removed them.

l.90 that fail L1 --> that otherwise fail L1.

Done

l91+3 the Fig.3 -> remove "the"

Done

section 4.3: you need to introduce the tracking 5 parameters - phi0 and t are not defined yet. This would fit well in the introduction of section 4, with two other sketches in rphi and rz planes.

Conclusion: please make entire sentences and no bullet points. The number of events collected should be quoted for the extended design, and then saying "which is more than one order of magnitude better than the current baseline tracking algorithm." or something along this line.

l.145 heavy Heavy -> heavy

Done

ARC comments (PAS v6 / AN v6)

comments from Andrei Gritsan

Title and abstract: do you want to spell out â?oL1â?�, â?oHLâ?�, mention CMS-II or something like that? You do spell them out in the Intro, but it might be also appropriate earlier.

Done

General: I think Phase-2 of CMS is not referenced or discussed, but you need that. Also, labels on the plots with â?oCMSâ?� may be confusing if it is Phase-1 or Phase-2. My general suggestion to all FTR notes is to adopt some convention like "CMS-II" or "CMS Phase-II,â?� whatever works best.

Done

Physics process: use either "\phi" or â?oaâ?� but consistently. Also, suggest to use "H(125)" instead of â?ohâ?� for the SM Higgs (check PDG, they do not use h. The h notation comes from our theory colleagues when they deal with 2HDM or something like that, but it makes it hard to follow when you have experimental results. H has been established experimentally as H, not as h, and we write experimental notes). You also use H(250) later, so H(125) will go along with that convention.

Done

Line 3: you write the number of expected H bosons (perhaps define H here), but why not quote lumi, energy, and pp collisions to define what we discuss in the rest of the note?

Done

Line 27: "Standard CMS samplesâ?� sounds like jargon to me. I assume these are Phase-II of CMS samples, which are not that standard for CMS as we do physics now. Connect lines 25-28 together, and describe and reference things better.

We removed the word "standard"

Line 33: It is not lifetime but c\tau if measured in cm.

Yes, you are right. Fixed it.

Line 37: need space between the sentences.

Done

Line 49: need comma after 0.3.

Done

Between lines 63-64: restore line numbers (e.g. change how to use equation, array vs regular one, but please fix). On the second line here and in many many other places you write â?oFig.â?� but it should be â?oFigureâ?� when used as the first word of a sentence. You can still use â?oFig." when in the middle of a sentence. There are very many cases to fix.

Done

Fig.1: x-axis label, legend, all look like collection of jargon. What is â?oGenâ?�, what is â?oTTBarâ?�. It is all CMS jargon. Please make this and all other plots professional-looking. Extend the y-axis range down to zero here. You mix up dark and light in the caption.

We explained the x-axis label in the caption. We fixed the dark and light mix up.

Fig.2: On this plot you put y-axis label in the middle of the plot, but x-axis label and labels on many other plots are in the Root style at the edge. Either way is fine, and I personally think having labels in the middle is better (it was a bad influence of Root to have it at the edge), but please make it consistent everywhere. I also think you can unify the style of the legend on all plots, the Root style box here is very different from everywhere else. You also use grid on this plot, but not on other plots.

We changed it The grid in Figure 2 is removed and the labels are no longer centered. We also removed the box for the legend.

Fig.4: label the z-axis (color coding) as efficiency for example.

We explained this in the caption of these images.

Conclusion: make a stronger conclusion where you actually state some conclusions. Just read the last sentence of the conclusion chapter and conclude that it is not really what you write as the concluding sentence.

We updated the conclusion, please take a look.

comments from author-ARC meeting on Sep 28, 2018

[1]:

- Add more explanation to improve the text for a better overall coherent story. Perhaps we stressed those two sections, but we also commented about Secs. 3 and 4, to tell a story about baseline vs. displaced track opportunities, and make text flow smooth overall.

We have updated the text, please take a look.

- Use either (phi) or (a).

Done. Using phi

- Figure 1: Fix the typo in caption (Dark Blue=> fast jet and light blue => 2 layer cluster)

Done

-Figures in general: Captions are small, could add more text to it, make them more descriptive.

We are added more description to the captions

-State trigger thresholds used in the captions.

We show the trigger thresholds as a function of the kinematic parameters.

-L1 rates plot: State that with displaced tracking we can trigger on signals with low HT, which was not possible previously.

Done.

-Examine the possibility to reweighs Powheg Pt to NNLO (This might not be needed)

This seems like more work than it's worth at this point.

- About the trigger, efficiency to see if we have an easy illustration of the following ratio in either physics reconstruction or object reconstruction. The ratio would be selection efficiency after trigger and offline reconstruction to selection efficiency after offline reconstruction only (no trigger involved, as if every single event of interest is triggered with 100% efficiency).

-add reference to phase 2 CMS detector

Done

-add description of phase 2 MC samples

Done

-add more emphasis or clarify that the displaced tracking is beyond the current baseline for the CMS phase 2 upgrades.

We have updated the text, please take a look

-make stronger claim about impressive/important prompt jets are before discussing displaced tracks.

We have updated the text, please take a look

-add trigger turn on curve plots to the AN

done

Comments From Anne-Marie Magnan

Section 3.1: can you give reasons / plots highlighting that these choices of sizes (0.2x0.23) in eta-phi - 6cm in z0 - at least 2(3) tracks depending on jets ET - are good ones ?

Table1: again, what drives the choices of the chi2 cuts ? Do I understand correctly from "the chi2 selections are per degree of freedom" that these numbers are chi2/NDF ?

l52-53 (and 60-61 in PAS) I do not understand this sentence: fake tracks distributed uniformely in 1/pT : what is distributed ? The fake rate ?

Could we have associated distributions ? Purity or fake rates as a function of track pT for example ?

The following Plot has the distribution of the genuine and non-genuine tracks as a function of 1/pt for all the tracks and tracks passing loose and tight selection.

* fake_pt.png:

l59 high compatibility: there is a factor 2 difference in rates.....

table 2 (and in PAS) can you specify < or > for the criteria ? And again, what drives the choice of cuts ? Is chi2 also the chi2/NDF here ?

The Loose Cuts are chosen individually, such that the cut value retains ~99% of the genuine tracks from the signal. Similarly, for the “Tight selection” we choose the cut values based on the Normalized distributions of the variables for fake and genuine tracks. And yes, It is Chi2/NDF.

* chi2_cuts.png:

l88 why the choice of d0>0.1 ?

|d0|<0.1 contains the core of the prompt trk d0 distribution. The plot below shows the do distribution for various lifetimes (Black: Minbias, Blue: ctau=0cm, Red ctau=1 cm, Green ctau=5cm and Yellow ctau=10cm)

* d0_trk.png:

Section 6 is empty....

l91-92 heavy higgs signal -> different signals considered

This has been rephrased.

Section 7: I guess this is my main concern, can we demonstrate somehow the independence of eta requirement extrapolation from prompt to displaced tracking algorithms ? Given the detectors have different geometries for barrel and endcap, are we really sure the extrapolation works ? I think it was discussed at the pre-approval already but I cannot remember the explanation/conclusion, sorry.

Appendix A: but Powheg is known to give better predictions, correct ? Given the impact on the results, is there a need to do a reweighting to NNLO+NNLL of the Higgs boson pT like is done in Run2 Higgs analyses ?

PAS- v6:

Abstract:

l1 track trigger vs l7 track triggers

Done

l6 scalar: a could also be a pseudoscalar, right ? The h->aa->4j should be described in words to make sure to explain all the bits, and a referred to as "a new light boson" to stay generic enough ?

updated it

Text:

l3 Higgs boson -> standard model (SM) Higgs bosons (h)

FIxed

l6 Something like "The case where the Higgs boson decays to two new light bosons, which in turn decay to jets, h->aa->4j, is of concern:..."

Done

scalar -> or pseudoscalar

We used scalars to generate the MC.

l16 jet lifetime tag for tracks with pseudorapidity |eta| < 1.0, otherwise the future work about extending to outer tracker cannot be understood.

Done

l21 Higgs boson decay to light (pseudo)scalars a bosons

Done

Before section 2: add a short paragraph about the CMS Phase 2 detector, with references to all TDRs.

To do

l25 GEANT -> GEANT4

Done

l27 200 -> this is Poisson 200, not exactly 200.....

Yes, it is Poisson 200

l31 impact parameter in the x-y plane (d0) ??

Yes.

l36 Hidden -> Heavy SM-like boson [I suppose this is a SM-like higgs boson, right ?) Do you mean "hidden" from the fact that we did not find it in the standard decay modes ? I think "hidden" is normally used to refer to hidden-sector theories, here I understand it's not the case....

done

l59 definition of chi2

We use chi^2 per NDF

l63+4 (if you add blank lines around equations it will avoid this problem of line numbers missing....) : HT and quadjet triggers->CMS jargon l65 definition of 11246 Hz, reference ?

To do

fig1 and 2: may want to introduce "TwoLayer jets" in text, or find it a better common name for everything......

We introduced it in the text.

l85 it is expected... see comment on AN.

Rephrased.

section 4.3: it could be useful to introduce the tracking 5 parameters - phi0 and t are not defined I think ?? Maybe this would fit well in the introduction of section 4, with two other sketches in rphi and rz planes ?

To do

l93 the loose selection is not used anywhere, is it ?

All the "displaced" tracks are required to pass loose selection and they are used to identify two-layer jet clusters.

Section 5: could do with an introduction giving specifications of the full trigger, commenting on the feasibility and explaining or introducing the benchmark points of 25 kHz , 5 kHz.

To do

l99-100 hidden Higgs boson signal -> Higgs boson signals in the different scenarios considered

We changed the hidden Higgs boson, to Heavy SM-like higgs boson

l113-114 confirmed that such extrapolation works -> isn't it kind of by construction though?

By confirming that we mean, We use prompt tracking to illustrate that this method this method works.

section 6: you cannot introduce new results in a conclusion. Move this in previous section, and conclusion should properly summarise the main points you want the reader to take from the note.

We updated the text please take a look.

l121-122 40 and 10 kHz has been replaced by 25 and 5 kHz, correct ?

Fixed it

Figure 9 missing unit for the x axis. Arb unit for the y-axis: is there a way this could be expressed "per fb-1 per pb" or some measurable quantity that could be extrapolated to different luminosity and signal cross section times BR scenarios [which would be valid only under SM-like Higgs assumptions but still could give useful quantitative information ?]

Fixed it

References:

Remove page range for [3] [6] [9]

Fixed

[4] not proper arXiv number

Fixed

[7] and [11] C. -> CMS

Fixed

[9] missing "o"Â due to special character I suppose.....

Fixed

--

PAS v4 Comments

Comments From Juliette Alimena

- abstract L2: “BSM” and “H_T” need to be defined.

- abstract L3: The symbols “h”, “a”, and “j” need to be defined.

- The abstract should mention the HL-LHC and/or Phase 2.

- L4: “BSM” needs to be defined.

- L4, 20: “Higgs boson decays”

- L4-5: The phrase “is getting events to pass the first level (L1) trigger” is a little colloquial and also implies that the L1 trigger can’t be changed or developed. Maybe “is designing a first level (L1) trigger algorithm that is suitable to collect these events”?

- L8: “Pile-up” needs to be defined. https://twiki.cern.ch/twiki/bin/view/CMS/Internal/PubGuidelines#Word_usage_and_jargon

- L8: “Fake” is a word that pubcomm asks us to avoid since it is vague (same reference as previous comment on this line).

- L9: “H_T” needs to be defined.

- L24: “Monte Carlo (MC) simulated events” https://twiki.cern.ch/twiki/bin/view/CMS/Internal/PubGuidelines#Word_usage_and_jargon

- L29: \pt needs to be defined as the transverse momentum, and \eta needs to be defined as the pseudorapidity.

- L30: \sigma needs to be defined as the width of the Gaussian distribution.

- L33: “H(250)” means that the Higgs mass you choose is 250 GeV? This should be written more clearly.

- Section 2: You should describe what generators are used to simulate these events.

- Section 2 is called “Signal and Background Samples” but only signal samples are mentioned in the text. Do you use any background samples?

- L39: “R parameter” —> “radius parameter”

- L45: The \phi symbol needs to be defined.

- L47: E_T needs to be defined.

- Figure 1, Figure 2 right, Figure 7: The x-axis needs units, presumably “[GeV]”.

- L67+: “Stub” needs to be defined.

- L67++: You have already defined d_0 as the impact parameter, no need to do it again.

- Figure 4 caption: No need to redefine \eta and d_0.

- Figure 4 caption: “as a Gaussian distribution with a width of 2 cm”

- L73: “the tracklet algorithm”? “a tracklet algorithm”? Furthermore, what is the tracklet algorithm?

- L82++++: The variables in this equation need to be defined.

- L87: Section 6 is empty? Or maybe Sections 7 and 8 were meant to be subsections of Section 6?

- L90: “Higgs boson signal” (capitalized Higgs and add “boson”)

- Figure 6 and 8 captions: “standard model Higgs bosons”, “heavy Higgs bosons” (“SM” has not been defined)

- Figure 6: mH and m\phi have not been defined, and their values presumably need GeV units.

- L101: “SF” has not been defined.

- L101-102: “in the full \eta range and in the central \eta range”

- L103: Missing figure reference.

- L104: \eta_J has not been defined.

- L105: The symbols in this ratio have not been defined.

- L107: “vs” —> “as a function of the”

- L107: “SM” has not been defined, so you need to write out “standard model” (or define it in the text, presumably earlier than this point)

- L108: Since the “Summary” contains additional information (namely, the money plot), you should instead call it “Conclusions”: https://twiki.cern.ch/twiki/bin/view/CMS/Internal/PubGuidelines#Section_headings

- L109-112: The symbols need to be redefined in the Summary/Conclusions section (h, a, j, H_T)

- I think a better concluding sentence would be useful.

Type A

- title: I would write out “beyond the standard model”. “Higgs” needs to be capitalized, and you need to refer to “Higgs bosons”.

- abstract L3: “long-lived” typo

- L6-7: “but the majority of the signal events do not get to the analysis because they fail the L1 trigger” —> “but the majority of the signal events fail the L1 trigger”.

- L29: Capitalize “Displaced”

- L38: Insert a comma after “In this section”

- L39, L56, Figure 1 and 2 captions: Use the anti-\kt macro.

- Table 1 caption: “of a 4-parameter track fit” —> “for a 4-parameter track fit”

- L62, Figure 3 caption: Use the \pt macro.

- L67+++: “figure” —> “Fig.”

- L91: “rate” should be lowercase

- L94: “Figure” —> “Fig.”

- Figure 7: You need a space between “Rate” and “(kHz)”. The “T” in H_T should be a subscript in the legend and in the x-axis (and H_T on the x-axis needs units, see the comment above). Please remove the box around the legend.

- L102: Remove “which”, otherwise the sentence is a fragment.

- L104: “First, the L1 rate”

- L105: “based on the ratio”.

- L109: Use the passive voice in the Summary/Conclusions section: https://twiki.cern.ch/twiki/bin/view/CMS/Internal/PubGuidelines#Passive_versus_active_voice_and

- L111: “Figure” —> “Fig.”

- L111: Insert a comma after “Fig.\,9”

- L129: I think you are missing a colon and a space, and it should be

“Geant4

A simulation toolkit”.

- All the figures need to use the CMS macros to write “CMS Simulation”, “14 TeV”, etc. A few of your figures already do this, but the “\sqrt{s}=14 TeV” is outdated, please see the new recommendation. The figures should also say “200 PU”. Most of this writing is too small, please make it bigger.

Comments From Nicola De Fillippis

Comments From Andreas Meyer

* Text is sometimes a bit jargonic, e.g. abstract "jetty BSM signals" "infamous blind spot" and assumes prior knowledge that I dont have. Suggest to rephrase / add in a more self-contained way, also in the introduction about the physics case and the foreseen hardware infrastructure.

* line 31/33 respectIVELY.

* line 38: resilient maybe better than resistant ?

* line 39: should we write DELTA R parameter (here and elsewhere)

* line 51 and 96: then -> than

* line 55: "comparably" means "similar performance" ?

* figures: increase fonts of labels and legends

* section 4: maybe I missed it, suggest to explain how the "special version" (line 73) of the tracklet algorithm can be implemented in the custom L1 hardware. Is there space or capacity foreseen already ? large enough FPGA ?

* line 89: explain how the rate estimates where calculated.

* line 90: is 25 kHz a realistic assumption for the available bandwidth ?

* line 109: studies => studied

-- RishiPatel - 2018-09-05