Monday, April 20, 2015

ATLAS: 2.5-sigma four-top-quark excess

ATLAS has posted a new preprint
Analysis of events with \(b\)-jets and a pair of leptons of the same charge in \(pp\)-collisions at \(\sqrt s = 8\TeV\) with the ATLAS detector
boasting numerous near-2-sigma excesses (which could be explained by vector-like quarks and chiral \(b'\) quarks, but are too small to deserve much space here) and a more intriguing 2.5-sigma excess in various final states with four top quarks.

This four-top excess is most clearly expressed in Figure 11.

This picture contains four graphs called (a),(b),(c),(d) – for ATLAS to celebrate the four letters of the alphabet ;-) – and they look as follows:

You may see that the solid black curve (which is sometimes the boundary of the red excluded region) sits strictly outside the yellow-green Brazil one-or-two-sigma band. The magnitude of the excess is about 2.5 sigma in all cases.

The excess is interpreted in four different ways. The graph (a) interprets the extra four-top events in terms of some contact interaction linking four stops at the same point. The horizontal axis shows the scale \(\Lambda\) of new physics from which this contact interaction arises. The vertical axis is the coefficient of the quartic interaction.

The graph (b) assumes that the four leptons come from the decay of two pair-produced sgluons whose mass is on the horizontal axis. On the vertical axis, there is some cross section times the branching ratio to four tops.

And the remaining graphs (c) and (d) assume that the four tops arise from two universal extra dimensions (2UED) of the real projective plane (RPP) geometry. The Kaluza-Klein mass scale is on the horizontal axis. The vertical axis depicts the cross section times the branching ratio again. The subgraphs (c) and (d) differ by using the tier \((1,1)\) and \((2,0)+(0,2)\), respectively.

Extra dimensions are cool but I still tend to bet that they will probably be too small and thus inaccessible to the LHC. Moreover, the RPP geometry is probably naive. But it's fun to see something that could be interpreted as positive evidence in favor of some extra dimensions.

I find the sgluons more realistic and truly exciting. They are colored scalar fields ("s" in "sgluon" stands for "scalar") in the adjoint representation of \(SU(3)_{QCD}\), much like gluons, and may be marketed as additional superpartners of gluinos under "another" supersymmetry in theories where the gauge bosons hide the extended, \(\NNN=2\) supersymmetry. Such models predict that the gluinos are Dirac particles, not just Majorana particles as they are in the normal \(\NNN=1\). This possibility has been discussed on this blog many times in recent years because I consider it elegant and clever – and naturally consistent with some aspects of the superstring model building.

Their graph (b) shows that sgluons may be as light as \(830\GeV\) or so.

Previously, CMS only saw a 1-sigma quadruple-top-quark "excess".

Finally, I also want to mention another preprint with light superpartners, ATLAS Z-peaked excess in MSSM with a light sbottom or stop, by Kobakhidze plus three pals which offers a possible explanation for the recent ATLAS Z-peaked 3-sigma excess. They envision something like an \(800\GeV\) gluino and a \(660\GeV\) sbottom.


  1. Hey, Lubos - this is off-topic, but I have a question for you and can't find a separate email address:

    I have gotten a request from the Director of this year's National Book Festival here in the U.S. for recommendations on books:

    - On "hard science" topics (physics!), but that might be interesting to the intelligent layman; and:

    - That have been published in the last two years or so;

    - The author of which would be willing to go to D.C. this September and give a presentation on the book.

    The last condition would be subject to negotiation, but we would value your ideas especially as to the first. Thanks.

  2. Would graphs (a) and (b) be evidence of supersymmetry?

  3. John F. HultquistApr 21, 2015, 6:10:00 AM

    I have to tell you this, even though it is silly.

    In the text, there is the phrase “The graph (a) interprets the extra four-top events ...”
    Then, at the end of the paragraph there is an inserted graphic ad for the 50th anniversary album of the group, the Four Tops.

    That is not AI at its best.

  4. Dear John, it is not an ad. I inserted the picture of four tops because the article is about four tops and it is witty.

    This is not artificial intelligence at its best because it's not artificial intelligence at all. It is real human intelligence and you lack all sense of humor.

  5. Dear Sup, (a) shows the (so far small) excess explained in a theory-neutral way. Even if the excess were more than a fluke, it could have many different explanations. SUSY (some forms of it) would be one of them. So the excess would be circumstantial evidence for SUSY - but not just SUSY.

    (b) shows pretty much the same excess (information about it) in terms of sgluons, and if the sgluons were there, supersymmetry would be found in Nature, too. But the excess in a similar graph "assuming" the existence of sgluons doesn't mean that that the assumption is the best explanation of the excess. It is *a* possible explanation and further tests would be needed to pick the best one.

  6. Apologies, I don't see an efficient way for me to help you with your work.

  7. Heh. Neither Latin nor Greek has any articles. Those were smart people.

    Indicative (Latin: [modus] indicativus), which states facts:

    The slave is carrying wine.
    servus vinum portat.

    Subjunctive or Conjunctive (Latin: [modus] conjunctivus), which is used for possibilities, intentions, necessities, and statements contrary to fact:

    May the slave carry the wine.
    servus vinum portet.

  8. Do you have any idea how to make sense of quantum mechanics in the case of de sitter space ? You don't have arbitrarily large measurement apparatus and you can't make an infinite number of experiments. I don't know what to do if I can't define classical degrees of freedom.

  9. Prathyush ManchalaApr 21, 2015, 12:19:00 PM

    Have you ever encountered actions with terms of this form.

    Or possibly Spinors Contracted with n surface area elements.


  10. You can Google for books physics 2014 2015 or some such keyword, then contact the authors.


  11. Dear John, one doesn't need any exactly "classical" degrees of freedom for quantum mechanics to make sense. For example, the Hilbert space of many spins (e.g. in a quantum computer) has no (or a trivial) classical counterpart.

    More generally, quantum mechanics doesn't depend on classical physics.

    What you may be worried about is that there isn't ever an infinite amount of decoherence for information to become classical. But that's true in all everyday situations, in or outside the de Sitter space, too. It just means that no observer's measurements or questions are ever completely accurately formulated and the answers are never perfectly exactly calculable.

  12. Ha! Your English writing is fantastic, far better tnan most native speakers, and not just because of the deep content, but also for the clever, original turns of phrase. There is only one Lubos and we readers are so lucky to have TRF.

  13. Dear Lubos, thank you very much for your answer. I wanted to mean impossibility of having classical information like you have said in your third paragraph. That is you have to have some kind of classical degrees of freedom (may be degrees of freedom is wrong word here, but result of every measurement is ultimately defined by some classical property of measuring apparatus ?) to be able to talk about measurement (if this is wrong correct me, I know you don´t need it in Consistent Histories, but to speak about experiments or to use copenhagen formalism you need it). Of course there quantum variables which don't have classical counterparts.

    Of course we never have infinite amount of decoherence but don´t we implicitly assume that there we can always make everything more precise buy using better apparatus. My worry is the possibility that there may be some theoretical limit on the amount of decoherence.

  14. kashyap vasavadaApr 21, 2015, 4:03:00 PM

    I agree completely with you.In fact I am delighted that all of this is pretty much identical with eastern religious philosophy.

  15. If I had the same request, I would definitely contact to Kip Thorne to give a lecture on his book "The Science of Interstellar". Success assured. Too expensive? Then the next stop would be surely Max Tegmark and his "Our Mathematical Universe: My Quest for the Ultimate Nature of Reality". And finally, if I get negatives in the previous two, a most modest bet coul be Jon Butterworth and his "Smashing Physics - Inside the World's biggest experiment". They are all recent books and apt for intelligent layman readers.

  16. kashyap vasavadaApr 21, 2015, 4:50:00 PM

    Hi Lubos: I have a question. I took QFT class in ancient times (1962!). In those days they used to teach QFT (especially perturbative QFT) using interaction picture, in between Schrodinger and Heisenberg pictures. I am not familiar about the latest trend. Do they use purely Heisenberg picture?

  17. Dear Kashyap, your question is a good one, but only for an undergraduate (with a very diplomatic instructor) - and/or highly outdated - one. And when I say "outdated", I mean 1926-outdated, not just 1962-outdated.

    The equivalence between the H. and S. (and Dirac) pictures has been understood since 1926, or the following years, and good undergraduate students have been able to understand the equivalence from the beginning, too.

    So they would agree that the question is silly.

    Moreover, the Dirac picture is only well-defined when there is a "natural" split of the Hamiltonian to the free part and the interacting part. In practice, it's only the case if the field theory is described as a perturbative expansion - usually an expansion around a free (Gaussian) theory.

    The Dirac picture is one of the most convenient ways to compute e.g. the S-matrix for such perturbative quantum field theories. But modern courses also use the Feynman path integral to achieve the same thing - the Feynman path integral formalism allows one to split the Lagrangian in the two ways in analogy with the splitting of the Hamiltonian in the Dirac picture.

    Good instructors don't have a problem in between them, and they choose the formalism slightly randomly. So you find courses based on the path integral as well as the interacting picture etc. The path integral is better for the treatment of the field theories that became very important after 1962 - gauge theories with gauge symmetries etc. So the path integral is preferred by modern courses in recent decades.

    However, the very act of looking at the field theories in the perturbative way is a bit outdated. People deal with nonperturbative phenomena and even field theories that aren't formulated as any perturbative expansion at all. So the Dirac picture becomes ill-defined for them. One accesses the full theory and its observables - by which I mean Green's functions or S-matrix here, not "operators" - directly, using its values and properties, or conformal conditions, or in other ways that depend on the context.

    Given all these issues, I think that your question only makes sense as a sociological one, and only within a very narrow context of "courses by instructors who don't really understand things well". I don't know what "picture" such folks would be choosing but it may be hard to become a good field theorist by being exposed to such lousy courses.

  18. kashyap vasavadaApr 21, 2015, 5:51:00 PM

    Thanks for clarification.My question was not about the equivalence. That I knew for sure. But it was about the preferred method. But as you say, path integrals would be preferred today. OK! I will ask around and find out if the instructors are ancient like me or modern!!

  19. Actually I'm by inclination a pretty materialist, reductionist sort of guy, and while I believe there are probably things that are beyond human intellectual comprehension, they are unfortunately also not accessible by any other "non-intellectual" means such as gnosticism or revelation. I am not antagonistic to religion which it seems to me can have tremendous social and ritualistic value for individuals and groups so inclined. The intellectual content of religion can be useful if it supports a positive ethical stance, and destructive otherwise.

  20. kashyap vasavadaApr 21, 2015, 6:16:00 PM

    One additional point. It seems that once you go beyond Harvard, MIT, Princeton, Stanford, Berkeley etc, European (eastern European?) education may be much ahead in mathematics and theoretical physics in at least undergraduate physics program than American system. I was talking to one of the younger faculty from eastern europe. After high school, people who go into physics, concentrate just on physics and math in Europe. But in American system they take lot of other courses like History, English, Philosophy, some other liberal arts courses not to mention Chemistry and Biology ! I think this discussion came up before, but I am not sure. Any way your point about modernization of QM courses is well taken.

  21. Ha, that's standard Lubos humour, I get confused by those regularly, too : )

  22. kashyap vasavadaApr 21, 2015, 6:24:00 PM

    Yes! It is a 64000$ question whether there is a way to find out about the universe by other than empirical science, whether there is an extra sensory world or not!

  23. Very interesting! I was a bit surprised to see the word "sgluon" at first, because I thought you meant "gluino", and then I realised that "s" is for scalar : )

  24. I suspect "really" was being ironic or sarcasitc or both. I read him to be a theist or agnostic parodying dogmatic atheistic hostility to any manifestation of religion in current intellectual discourse.

  25. I was also certain that Slavic languages have no definitive articles, but recently I learned about a counter-example: Bulgarian. The location of this definitive article is also a bit odd, viz. it is postfixed to the noun.

  26. Dear Pavle, I was just listening (again) a fun story about this Bulgarian exception yesterday:

    They analyze Kollár's "Slávy dcera" (Daughter of the Slavic Motherland, sort of), and there is "zem ta", meaning "the country", and Mládek says that it's normal in Bulgarian.

    For example, the guy was in a Bulgarian shop. And what did he say to get what he needed? Bitte una guma ta und una viagra ta. :-)

  27. John F. HultquistApr 22, 2015, 6:37:00 AM

    Oh, sorry. I didn't realize you had a
    sense of humor.

    Actually, I do.

    Anyway, I get all kinds of ads pushed
    at me and almost always ignore them. Now I will have to make a point
    of looking at the graphics on trf.

  28. Dear John, don't take me wrong. De Sitter space is hard and e.g. in string theory, we have much less clear equations governing such spaces than e.g. for anti de Siter spaces. SUSY has to be broken which makes things hard, and there is no asymptotic region that would allow a holographic description (sorry Andy Strominger with your dS/CFT).

    But I don't think that your assignment of the "burden of proof" is right.

    First of all, decoherence is never complete - in de Sitter space or otherwise. This is easy to see. The off-diagonal elements in the bases discussed by decoherence decrease quickly but they never go "strictly" to zero.

    This doesn't make quantum mechanics meaningless, or anything of the sort you are suggesting. It just means that we can't ask completely exactly formulated questions whose answers would be "classical" truth values.

    I need to emphasize that the word "classical" in the previous sentence in no way requires to use a "parallel" theory (classical physics). The word "classical" simply means "mutually commuting". It is a property of a pair of operators that is defined purely in the quantum formalism. If you want to assign particular truth values (answers) to two Yes/No questions, the operators representing these two questions have to be commuting because the numbers 0,1 are. This is a trivial fact and while "commuting operators" are similar to the quantities that existed in classical physics, one doesn't need any "classical theory" to be used alongside with quantum mechanics.

    If we can't ask the exact questions (find well-defined mutually exactly commuting operators etc.), no physical theory is obliged to calculate exact answers for these non-existent exact questions!

    |Psi|^2 is always interpreted as the probability or probability density etc. of something and of some sort. What it is the probability "of" may always be described in terms of operators etc. and one must be careful not to use the classical language and classical logic if two operators don't commute.

    But it is in no way guaranteed that there exists an apparatus that allows one to measure such quantities and probabilities arbitrarily accurately. It's common sense. Gadgets are never 100% accurate, they have never been, not even in classical physics. It is utterly irrational to claim that this trivial fact implies some problem with the theory.

    Your claims about the "unusual place of quantum mechanics" are nothing else than the standard idiotic anti-quantum zealotry. Quantum mechanics doesn't require its limiting case to be formulated. In fact, many important novel quantum theories don't have any classical limit, like the (2,0) theory at the conformal point.

    The only correct claim that may be obtained by fixing your wrong claim is the tautology that in order to describe phenomena in QM using a classical language, the objects in the classical language have to approximately obey the defining rule of classical physics - in particular, they have to commute with each other. If they don't commute, they can't simultaneously be assigned c-values. This is not a problem or incompeteness or anything else that would justify any negative word whatsoever, and whoever thinks otherwise is an anti-quantum zealot.

  29. Hi John,

    how could you miss Lumo s immensly funny sense of humor for so long ... ;-)?

    He often makes me rolling on the floor, spitting my coffee at my screen etc...
    And all this often in the midst of reading serious physics texts, LOL :-D

  30. Dear Lubos, thanks for you reply I really appreciate the time you spent.

    However your comment surprised me. Be sure that I understood Consistent Histories at least formally and know what are physical properties we can talk about. I derived rules of copenhagen formalism considering a system of an object with a macroscopic system. Some macroscopic properties of apparatus commute with each other and classical logic arises between these properties. This properties are actually what we see in our everyday life. In practice this properties only apprpximately commuting. I understand this and there is nothing wrong about this.

    However I have learned about these things only because you have written