Massive Gravity, or You Only Live Twice

Proving Einstein wrong is the ultimate ambition of every crackpot and physicist alike. In particular, Einstein's theory of gravitation -  the general relativity -  has been a victim of constant harassment. That is to say, it is trivial to modify gravity at large energies (short distances), for example by embedding it in string theory, but it is notoriously difficult to change its long distance behavior. At the same time, motivations to keep trying go beyond intellectual gymnastics. For example, the accelerated expansion of the universe may be a manifestation of modified gravity (rather than of a small cosmological constant).   

In Einstein's general relativity, gravitational interactions are mediated by a massless spin-2 particle - the so-called graviton. This is what gives it its hallmark properties: the long range and the universality. One obvious way to screw with Einstein is to add mass to the graviton, as entertained already in 1939 by Fierz and Pauli. The Particle Data Group quotes the constraint m ≤ 6*10^−32 eV, so we are talking about the De Broglie wavelength comparable to the size of the observable universe. Yet even that teeny mass may cause massive troubles. In 1970 the Fierz-Pauli theory was killed by the van Dam-Veltman-Zakharov (vDVZ) discontinuity. The problem stems from the fact that a massive spin-2 particle has 5 polarization states (0,±1,±2) unlike a massless one which has only two (±2). It turns out that the polarization-0 state couples to matter with the similar strength as the usual polarization ±2 modes, even in the limit where the mass goes to zero, and thus mediates an additional force which differs from the usual gravity. One finds that, in massive gravity, light bending would be 25% smaller, in conflict with the very precise observations of stars' deflection around the Sun. vDV concluded that "the graviton has rigorously zero mass". Dead for the first time...           

The second coming was heralded soon after by Vainshtein, who noticed that the troublesome polarization-0 mode can be shut off in the proximity of stars and planets. This can happen in the presence of graviton self-interactions of a certain type. Technically, what happens is that the polarization-0 mode develops a background value around massive sources which, through the derivative self-interactions, renormalizes its kinetic term and effectively diminishes its interaction strength with matter. See here for a nice review and more technical details. Thanks to the Vainshtein mechanism, the usual predictions of general relativity are recovered around large massive source, which is exactly where we can best measure gravitational effects. The possible self-interactions leading a healthy theory without ghosts have been classified, and go under the name of the dRGT massive gravity.

There is however one inevitable consequence of the Vainshtein mechanism. The graviton self-interaction strength grows with energy, and at some point becomes inconsistent with the unitarity limits that every quantum theory should obey. This means that massive gravity is necessarily an effective theory with a limited validity range and has to be replaced by a more fundamental theory at some cutoff scale 𝞚. This is of course nothing new for gravity: the usual Einstein gravity is also an effective theory valid at most up to the Planck scale MPl～10^19 GeV.  But for massive gravity the cutoff depends on the graviton mass and is much smaller for realistic theories. At best,

So the massive gravity theory in its usual form cannot be used at distance scales shorter than ～300 km. For particle physicists that would be a disaster, but for cosmologists this is fine, as one can still predict the behavior of galaxies, stars, and planets. While the theory certainly cannot be used to describe the results of table top experiments,  it is relevant for the  movement of celestial bodies in the Solar System. Indeed, lunar laser ranging experiments or precision studies of Jupiter's orbit are interesting probes of the graviton mass.

Now comes the latest twist in the story. Some time ago this paper showed that not everything is allowed  in effective theories.  Assuming the full theory is unitary, causal and local implies non-trivial constraints on the possible interactions in the low-energy effective theory. These techniques are suitable to constrain, via dispersion relations, derivative interactions of the kind required by the Vainshtein mechanism. Applying them to the dRGT gravity one finds that it is inconsistent to assume the theory is valid all the way up to 𝞚max. Instead, it must be replaced by a more fundamental theory already at a much lower cutoff scale,  parameterized as 𝞚 = g*^1/3 𝞚max (the parameter g* is interpreted as the coupling strength of the more fundamental theory). The allowed parameter space in the g*-m plane is showed in this plot:

Massive gravity must live in the lower left corner, outside the gray area  excluded theoretically  and where the graviton mass satisfies the experimental upper limit m～10^−32 eV. This implies g* ≼ 10^-10, and thus the validity range of the theory is some 3 order of magnitude lower than 𝞚max. In other words, massive gravity is not a consistent effective theory at distance scales below ～1 million km, and thus cannot be used to describe the motion of falling apples, GPS satellites or even the Moon. In this sense, it's not much of a competition to, say, Newton. Dead for the second time.   

Is this the end of the story? For the third coming we would need a more general theory with additional light particles beyond the massive graviton, which is consistent theoretically in a larger energy range, realizes the Vainshtein mechanism, and is in agreement with the current experimental observations. This is hard but not impossible to imagine. Whatever the outcome, what I like in this story is the role of theory in driving the progress, which is rarely seen these days. In the process, we have understood a lot of interesting physics whose relevance goes well beyond one specific theory. So the trip was certainly worth it, even if we find ourselves back at the departure point.

Per kaons ad astra

NA62 is a precision experiment at CERN. From their name you wouldn't suspect that they're doing anything noteworthy: the collaboration was running in the contest for the most unimaginative name, only narrowly losing to CMS...  NA62 employs an intense beam of charged kaons to search for the very rare decay K+ → 𝝿+ 𝜈 𝜈. The Standard Model predicts the branching fraction BR(K+ → 𝝿+ 𝜈 𝜈) = 8.4x10^-11 with a small, 10% theoretical uncertainty (precious stuff in the flavor business). The previous measurement by the BNL-E949 experiment reported BR(K+ → 𝝿+ 𝜈 𝜈) = (1.7 ± 1.1)x10^-10, consistent with the Standard Model, but still  leaving room for large deviations.  NA62 is expected to pinpoint the decay and measure the branching fraction with a 10% accuracy, thus severely constraining new physics contributions. The wires, pipes, and gory details of the analysis  were nicely summarized by Tommaso. Let me jump directly to explaining what is it good for from the theory point of view.

To this end it is useful to adopt the effective theory perspective. At a more fundamental level, the decay occurs due to the strange quark inside the kaon undergoing the transformation  sbar → dbar 𝜈 𝜈bar. In the Standard Model, the amplitude for that process is dominated by one-loop diagrams with W/Z bosons and heavy quarks. But kaons live at low energies and do not really see the fine details of the loop amplitude. Instead, they effectively see the 4-fermion contact interaction:

The mass scale suppressing this interaction is quite large, more than 1000 times larger than the W boson mass, which is due to the loop factor and small CKM matrix elements entering the amplitude. The strong suppression is the reason why the K+ → 𝝿+ 𝜈 𝜈  decay is so rare in the first place. The corollary is that even a small new physics effect inducing that effective interaction may dramatically change the branching fraction. Even a particle with a mass as large as 1 PeV coupled to the quarks and leptons with order one strength could produce an observable shift of the decay rate.  In this sense, NA62 is a microscope probing physics down to 10^-20 cm  distances, or up to PeV energies, well beyond the reach of the LHC or other colliders in this century. If the new particle is lighter, say order TeV mass, NA62 can be sensitive to a tiny milli-coupling of that particle to quarks and leptons.

So, from a model-independent perspective, the advantages  of studying the K+ → 𝝿+ 𝜈 𝜈  decay are quite clear. A less trivial question is what can the future NA62 measurements teach us about our cherished models of new physics. One interesting application is in the industry of explaining the apparent violation of lepton flavor universality in B → K l+ l-, and B → D l 𝜈 decays. Those anomalies involve the 3rd generation bottom quark, thus a priori they do not need to have anything to do with kaon decays. However, many of the existing models introduce flavor symmetries controlling the couplings of the new particles to matter (instead of just ad-hoc interactions to address the anomalies). The flavor symmetries may then relate the couplings of different quark generations, and thus predict  correlations between new physics contributions to B meson and to kaon decays. One nice example is illustrated in this plot:

The observable RD(*) parametrizes the preference for B → D 𝜏 𝜈 over similar decays with electrons and muon, and its measurement by the BaBar collaboration deviates from the Standard Model prediction by roughly 3 sigma. The plot shows that, in a model based on U(2)xU(2) flavor symmetry, a significant contribution to RD(*) generically implies a large enhancement of BR(K+ → 𝝿+ 𝜈 𝜈), unless the model parameters are tuned to avoid that.  The anomalies in the B → K(*) 𝜇 𝜇 decays can also be correlated with large effects in K+ → 𝝿+ 𝜈 𝜈, see here for an example. Finally, in the presence of new light invisible particles, such as axions, the NA62 observations can be polluted by exotic decay channels, such as e.g.  K+ → axion 𝝿+.

The  K+ → 𝝿+ 𝜈 𝜈 decay is by no means the magic bullet that will inevitably break the Standard Model.  It should be seen as one piece of a larger puzzle that may or may not provide crucial hints about new physics. For the moment, NA62 has analyzed only a small batch of data collected in 2016, and their error bars are still larger than those of BNL-E949. That should change soon when the 2017  dataset is analyzed. More data will be acquired this year, with 20 signal events expected  before the long LHC shutdown. Simultaneously, another experiment called KOTO studies an even more rare process where neutral kaons undergo the CP-violating decay KL → 𝝿0 𝜈 𝜈,  which probes the imaginary part of the effective operator written above. As I wrote recently, my feeling is that low-energy precision experiments are currently our best hope for a better understanding of fundamental interactions, and I'm glad to see a good pace of progress on this front.

Singularity is now

Artificial intelligence (AI) is entering into our lives.  It's been 20 years now since the watershed moment of Deep Blue versus Garry Kasparov.  Today, people study the games of AlphaGo against itself to get a glimpse of what a superior intelligence would be like. But at the same time AI is getting better in copying human behavior.  Many Apple users have got emotionally attached to Siri. Computers have not only learnt  to drive cars, but also not to slow down when a pedestrian is crossing the road. The progress is very well visible to the bloggers community. Bots commenting under my posts have evolved well past !!!buy!!!viagra!!!cialis!!!hot!!!naked!!!  sort of thing. Now they refer to the topic of the post, drop an informed comment, an interesting remark,  or a relevant question, before pasting a link to a revenge porn website. Sometimes it's really a pity to delete those comments, as they can be more to-the-point than those written by human readers.   

AI is also entering the field of science at an accelerated pace, and particle physics is as usual in the avant-garde. It's not a secret that physics analyses for the LHC papers (even if finally signed by 1000s of humans) are in reality performed by neural networks, which are just beefed up versions of Alexa developed at CERN. The hottest topic in high-energy physics experiment is now machine learning,  where computers teach  humans the optimal way of clustering jets, or telling quarks from gluons. The question is when, not if, AI will become sophisticated enough to perform a creative work of theoreticians. 

It seems that the answer is now.

Some of you might have noticed a certain Alan Irvine, affiliated with the Los Alamos National Laboratory, regularly posting on arXiv single-author theoretical papers on fashionable topics such as the ATLAS diphoton excess, LHCb B-meson anomalies, DAMPE spectral feature, etc. Many of us have received emails from this author requesting citations. Recently I got one myself; it seemed overly polite, but otherwise it didn't differ in relevance or substance from other similar requests. During the last two and half years,  A. Irvine has accumulated a decent h-factor of 18.  His papers have been submitted to prestigious journals in the field, such as the PRL, JHEP, or PRD, and some of them were even accepted after revisions. The scandal broke out a week ago when a JHEP editor noticed that the extensive revision, together with a long cover letter, was submitted within 10 seconds from receiving the referee's comments. Upon investigation, it turned out that A. Irvine never worked in Los Alamos, nobody in the field has ever met him in person, and the IP from which the paper was submitted was that of the well-known Ragnarok Thor server. A closer analysis of his past papers showed that, although linguistically and logically correct, they were merely a compilation of equations and text from the previous literature without any original addition. 

Incidentally, arXiv administrators have been aware that, since a few years, all source files in daily hep-ph listings were downloaded for an unknown purpose by automated bots. When you have excluded the impossible, whatever remains, however improbable, must be the truth. There is no doubt that A. Irvine is an AI bot, that was trained on the real hep-ph input to produce genuinely-looking  particle theory papers.     

The works of A. Irvine have been quietly removed from arXiv and journals, but difficult questions remain. What was the purpose of it? Was it a spoof? A parody? A social experiment? A Facebook research project? A Russian provocation?  And how could it pass unnoticed for so long within  the theoretical particle community?  What's most troubling is that, if there was one, there can easily be more. Which other papers on arXiv are written by AI? How can we recognize them?  Should we even try, or maybe the dam is already broken and we have to accept the inevitable?  Is Résonaances written by a real person? How can you be sure that you are real?

Update: obviously, this post is an April Fools' prank. It is absolutely unthinkable that the creative process of writing modern particle theory papers can ever be automatized. Also, the neural network referred to in the LHC papers is nothing like Alexa; it's simply a codename for PhD students.  Finally, I assure you that Résonaances is written by a hum 00105e0 e6b0 343b 9c74 0804 e7bc 0804 e7d5 0804 [core dump]