Type B
Title: "new background prediction seems to us quite unspecific, maybe you could add "using sidebands in data" or something like that?

Abstract:
The first sentence of the abstract reads a little odd, it might work better if it were reordered, or split into two (searches ....are presented. Proton-proton collision data ... of 137 /fb is used.)

Abstract.4: "using two complementary data-driven methods" -> to match the title, you should mention here that one of these two is a new method

Abstract.10: "extending previous searches" refers to all three signal scenarios here, right? Then you should put a comma before this phrase.

L 17: It is strange to mention "the previous CMS search" at this point when the full set of previous searches is only described at the end of the paragraph. We think it would be better for the reader if not referred to this single search, but instead say which parameters are used.

L52: ".. are used for the high-mass search" - This is a bit vague: what is the high-mass search in this case? One could argue the entire search is a high mass search, explicitly writing the mass range for which this trigger is used might be better.

L 57: Are charged leptons and pfMET used later on? If not, they could be omitted. If charged leptons are important, they should be defined, i.e. which pt threshold and ID. We actually think this is necessary only for the muon that is used to measure the trigger eff.

L83: In L79 it was already stated that jets are required to have |eta|<2.5. In this line we see "For PF-jets that satisfy |eta|<2.5...", which seems redundant. Is the |eta|<2.5 in L79 a typo? If not, then the explicit eta range is not required in the discussion at L83. Suggestion: Replace "For PF-jets that satisfy |η| < 2.5," by "Since all jets are required to be"

L88: "previous CMS searches [16, 19, 20, 22, 23, 25, 26, 29, 43]" -> it should be enough to mention just the one or two most recent results here, or the one where this specific analysis part is discussed in greater detail. Suggestion: Write ... as introduced in [16].

L103 "CR high is defined such that it has four to five times more background events than the SR" -> The way it is written now it reads as if the cut that defines the CR is specifically determined such that the CR has 4-5x more background events than the SR. Is this the case, or was the control region defined based on some other criteria, which result in the end in having 4-5x as many events as in the SR? If the latter it might make sense to rephrase to reflect that (for example by removing "is defined such that it")

L109: What is the difference between "turning on quickly" and "rapidly plateauing at 100%"? One of the two statements seems redundant (suggest to remove the 2nd)

L133ff: The discussion at the end of section 4, about the "special" event seems a bit out of place. The observed limits in the high mass regime barely show a 1 sigma deviation from the expected. One could argue that a search for a pair of di-jet resonances should be performed, but even in that case: this is one single event, and, with the background present, not statistically significant. Nevertheless we think it is nice to show an event display in the paper. If it was discussed in the theory community, triggered by the earlier PAS, we wonder if there is more than one paper to be cited? Is [48] the only one?

L143: If one does not look up the citation, one thinks this is a link to another CMS paper where this event was already discussed. This could e.g. be avoided by adding:
recently explored -> recently phenomenologically explored

L144: An additional paragraph explaining when which method is used should be added here. You could also explain here that the ratio method works only for masses > 2.6 TeV.

L152  "Symbolically..." Omit this sentence since indeed the following lines explain the method not only symbolically.

L160: Quote version number for POWHEG

L161: Instead of "jet η's" "jet pseudorapidities" might be better

L.165: The CR are not introduced here, but were already before: "... way, introducing a second ..." -> "... way from the second..." (and comma after CT_middle)

L 166: utilizing a |∆η| sideband as described previously. -> utilizing the |∆η| sideband described in Section 3.

Eq (5): Is the 3rd power necessary? Maybe a square would have done as well? How did you optimize this?

L180: "The systematic uncertainty ... freely floating ones" -> Profiling just means the parameters are freely floating in the fit. This has already been mentioned a few lines before

Fig. 4: Phrases as: "...corrections added on the top” are somehow misleading. Better: R_aux. auxiliary transfer factor for data (black points), PYTHIA (blue line), and POWHEG with electroweak corrections (top)  along (bottom) with their ratio ...
BTW It would be easier to understand if one calls this a double ratio. Similar an axis label ( “quantity/PYTHIA” ) Raux^{data}/Raux^{Simulation} would be better.
 
L207: "the same as those used by previous ... searches ... performed by the CMS collaboration" -> shouldn't there be one (the main) reference here?

L218: What is done with the other signal processes when a limit is set for example for qq? Are they fixed to 0 or also floating freely in the fit? Would be useful to state the assumption.

L219: "Dominant sources of systematic uncertainties: -> are other uncertainties applied, e.g. due to the finite size of the simulated signal samples? Or are the signals parameterised with a function? If so this wasn't clear up to now, but maybe we missed it.

L221: The uncertainty on the JES is eta and pT dependent, and most likely not the same for all three years, but you don't comment at all about it. If 2% is the average, it would be good to state this, and also expand a bit about the other dependences, since this is the main systematic, and this section is anyhow small. While, e.g. for lumi you state the uncertainty for the three years separately.

L231: "The modified frequentist ... resonance mass increases" This is a very long description that doesn't need to be this long. What is meant by "evaluated independently"? It's confusing because the background shape is very much correlated between the different bins. At the end, profiling is mentioned for the third time.

L252 What is the added value of reporting the local significances in the paper? From the limits one can already deduce that they are not going to be large...


Type A:

Abstract.8: "of narrow resonances, and broad spin-1" -> no "and" between the first and second item of a three-item list

Abstract.11: "and exclude, at 95% confidence level" -> no comma here

L3: "String resonances [1,2]" has too much space before the opening bracket. I see in your texcode "resonances ~\cite" -> should be "resonances~\cite" without an additional space. This happens at a lot of places in the paper.

L 23: Run II -> Run 2

L40 "where these latter detectors" -> I think 'latter' can only ever refer to the last thing, not the last two. To reduce ambiguity might be better to place a full stop after "calorimeters". And then write "The calorimeters consist of ..."

L51: Check if offline is jargon and replace if so

L64-66: This can be shortened and made easier to read, for example by rephrasing as "Triggers that require HT, the scalar sum of jet pT for all jets in the event with |eta|<3.0, to exceed a threshold are also used" or similar.

L66-69: This is a very long and hard to read sentence. Check if it can be broken up.

L100: Since CRmiddle denotes one control region, the acronym ' control regions (CR)' should be defined as ' control regions (CRs)' and later where multiple control regions are discussed at the same time, e.g. 'the two CR', 'CRs' should be used.

L100-103: very long sentence, can it be split up to make easier to read?

L101: Constraint -> Constrain

L103: CR high is to -> CR high is used to

L112: This is why... -> Therefore, ...

L129: Extra space between "Figure' and '2'

L130: "falls"->"fall" (because of dijet mass spectra on L129)

L138: has transverse momentum, pT, 3.5 TeV -> has a pT of 3.5 TeV

Fig1: y-axis: why not simply "Nevts with a  'x10^3' next to the axis?"

Fig. 2: legend: use "PYTHIA" as in the text
           legend: increase the spacing between the three lines "CR_high", "CR_middle", "SR"
           caption: it should "boxes" not "crosses" 

L151: is ratio -> is given by the ratio

L152: in SR -> in the SR

Eq. (3): add comma after equation, and then continue with "where" (and no comma needed after where)

L 157: "missing higher orders from LO Pythia simulation" -> "higher orders missing from LO Pythia simulation"

L164: "mostly at the high dijet mass region" -> "mostly at high dijet mass" or "mostly in the high dijet mass region"

168: R_aux shown -> add comma: R_aux, shown

L172: "separatelly" -> "separately"

L173: "are shown in Fig.4 bottom panels" -> "are shown in the bottom panels of Fig. 4"

L184: "model.For" -> "model. For"

Fig. 4, legend: Use consistently "POWHEG" in all captions (it's lower case in the lower left panel)

L191: predictions methods -> prediction methods

L195: method.This -> method. This (add space)

Fig5 caption: "is being used" -> "is used"

L235: in 100-GeV steps. -> in steps of 100 GeV

L238: , which as we descrived previously is automatically evaluated via profiling, -> , which, as we described previously, is automatically evaluated via profiling,  (two commas and one typo correction)

L 241: Figures 6 -7 -> Figures 6 and 7

L247, L258: write either "sigma x B x A" or "sigma B A" consistently in all places

L253:  in 100-GeV steps. -> in steps of 100 GeV

L271: don't break "95% CL"

Fig. 7, legend: uniform font size for different model names

L294: shapes.The -> shapes. The (add space)
L 304 : "using using" -> "using"

L308: "fit method.No evidence" -> missing space

L316: "limits limits" -> "limits"

Ref. [52], [53], [54]: Use the correct title which is different for each of the luminosity PASes, and the correct years