arXiv subjects

John Kearney

Publications and source records attributed to John Kearney.

16 recordsLinked to original sources

arXiv2019

Active Learning for New Domains in Natural Language Understanding

We explore active learning (AL) for improving the accuracy of new domains in a natural language understanding (NLU) system. We propose an algorithm called Majority-CRF that uses an ensemble of classification models to guide the selection of relevant utterances, as well as a sequence labeling model to help prioritize informative examples. Experiments with three domains show that Majority-CRF achieves 6.6%-9% relative error rate reduction compared to random sampling with the same annotation budget, and statistically significant improvements compared to other AL approaches. Additionally, case studies with human-in-the-loop AL on six new domains show 4.6%-9% improvement on an existing NLU system.

cs.CL↗

arXiv2017

Higgs-photon resonances

We study models that produce a Higgs boson plus photon ($h^0 γ$) resonance at the LHC. When the resonance is a $Z'$ boson, decays to $h^0 γ$ occur at one loop. If the $Z'$ boson couples at tree-level to quarks, then the $h^0 γ$ branching fraction is typically of order $10^{-5}$ or smaller. Nevertheless, there are models that would allow the observation of $Z' \to h^0 γ$ at $\sqrt{s} = 13$ TeV with a cross section times branching fraction larger than 1 fb for a $Z'$ mass in the 200--450 GeV range, and larger than 0.1 fb for a mass up to 800 GeV. The 1-loop decay of the $Z'$ into lepton pairs competes with $h^0 γ$, even if the $Z'$ couplings to leptons vanish at tree level. We also present a model in which a $Z'$ boson decays into a Higgs boson and a pair of collimated photons, mimicking an $h^0 γ$ resonance. In this model, the $h^0 γ$ resonance search would be the discovery mode for a $Z'$ as heavy as 2 TeV. When the resonance is a scalar, although decay to $h^0 γ$ is forbidden by angular momentum conservation, the $h^0$ plus collimated photons channel is allowed. We comment on prospects of observing an $h^0 γ$ resonance through different Higgs decays, on constraints from related searches, and on models where $h^0$ is replaced by a nonstandard Higgs boson.

hep-ph↗

arXiv2017

$Z$ boson mediated dark matter beyond the effective theory

Direct detection bounds are beginning to constrain a very simple model of weakly interacting dark matter---a Majorana fermion with a coupling to the $Z$ boson. In a particularly straightforward gauge-invariant realization, this coupling is introduced via a higher-dimensional operator. While attractive in its simplicity, this model generically induces a large $ρ$ parameter. An ultraviolet completion that avoids an overly large contribution to $ρ$ is the singlet-doublet model. We revisit this model, focusing on the Higgs blind spot region of parameter space where spin-independent interactions are absent. This model successfully reproduces dark matter with direct detection mediated by the $Z$ boson, but whose cosmology may depend on additional couplings and states. Future direct detection experiments should effectively probe a significant portion of this parameter space, aside from a small coannihilating region. As such, $Z$-mediated thermal dark matter as realized in the singlet-doublet model represents an interesting target for future searches.

hep-ph↗

arXiv2017

Spacetime Dynamics of a Higgs Vacuum Instability During Inflation

A remarkable prediction of the Standard Model is that, in the absence of corrections lifting the energy density, the Higgs potential becomes negative at large field values. If the Higgs field samples this part of the potential during inflation, the negative energy density may locally destabilize the spacetime. We use numerical simulations of the Einstein equations to study the evolution of inflation-induced Higgs fluctuations as they grow towards the true (negative-energy) minimum. These simulations show that forming a single patch of true vacuum in our past light cone during inflation is incompatible with the existence of our Universe; the boundary of the true vacuum region grows outward in a causally disconnected manner from the crunching interior, which forms a black hole. We also find that these black hole horizons may be arbitrarily elongated---even forming black strings---in violation of the hoop conjecture. By extending the numerical solution of the Fokker-Planck equation to the exponentially suppressed tails of the field distribution at large field values, we derive a rigorous correlation between a future measurement of the tensor-to-scalar ratio and the scale at which the Higgs potential must receive stabilizing corrections in order for the Universe to have survived inflation until today.

hep-ph↗

arXiv2016

Multi-step production of a diphoton resonance

Among the questions that would be raised by the observation of a new resonance at the LHC, particularly pressing are those concerning the production mechanism: What is the initial state? Is the resonance produced independently or in association with other particles? Here we present two weakly-coupled renormalizable models for production of a diphoton resonance that differ in both their initial and final states. In one model, a scalar particle produced through gluon fusion decays into a diphoton particle and a light, long-lived pseudoscalar. In another model, a $Z'$ boson produced from the annihilation of a strange-antistrange quark pair undergoes a cascade decay that leads to a diphoton particle and two sterile neutrinos. Various kinematic distributions may differentiate these models from the canonical model where a diphoton particle is directly produced in gluon fusion.

hep-ph↗

arXiv2016

High-Scale Axions without Isocurvature from Inflationary Dynamics

Observable primordial tensor modes in the cosmic microwave background (CMB) would point to a high scale of inflation $H_{I}$. If the scale of Peccei-Quinn (PQ) breaking $f_a$ is greater than $\frac{H_{I}}{2π}$, CMB constraints on isocurvature naively rule out QCD axion dark matter. This assumes the potential of the axion is unmodified during inflation. We revisit models where inflationary dynamics modify the axion potential and discuss how isocurvature bounds can be relaxed. We find that models that rely solely on a larger PQ-breaking scale during inflation $f_I$ require either late-time dilution of the axion abundance or highly super-Planckian $f_I$ that somehow does not dominate the inflationary energy density. Models that have enhanced explicit breaking of the PQ symmetry during inflation may allow $f_a$ close to the Planck scale. Avoiding disruption of inflationary dynamics provides important limits on the parameter space.

hep-ph↗

arXiv2016

Tadpole-Induced Electroweak Symmetry Breaking and pNGB Higgs Models

We investigate induced electroweak symmetry breaking (EWSB) in models in which the Higgs is a pseudo-Nambu-Goldstone boson (pNGB). In pNGB Higgs models, Higgs properties and precision electroweak measurements imply a hierarchy between the EWSB and global symmetry-breaking scales, $v_H \ll f_H$. When the pNGB potential is generated radiatively, this hierarchy requires fine-tuning to a degree of at least $\sim v_H^2/f_H^2$. We show that if Higgs EWSB is induced by a tadpole arising from an auxiliary sector at scale $f_Σ\ll v_H$, this tuning is significantly ameliorated or can even be removed. We present explicit examples both in Twin Higgs models and in Composite Higgs models based on $SO(5)/SO(4)$. For the Twin case, the result is a fully natural model with $f_H \sim 1$ TeV and the lightest colored top partners at 2 TeV. These models also have an appealing mechanism to generate the scales of the auxiliary sector and Higgs EWSB directly from the scale $f_H$, with a natural hierarchy $f_Σ\ll v_H \ll f_H \sim{\rm TeV}$. The framework predicts modified Higgs coupling as well as new Higgs and vector states at LHC13.

hep-ph↗

arXiv2016

Is a Higgs Vacuum Instability Fatal for High-Scale Inflation?

We study the inflationary evolution of a scalar field $h$ with an unstable potential for the case where the Hubble parameter $H$ during inflation is larger than the instability scale $Λ_I$ of the potential. Quantum fluctuations in the field of size $δh \sim \frac{H}{2 π}$ imply that the unstable part of the potential is sampled during inflation. We investigate the evolution of these fluctuations to the unstable regime, and in particular whether they generate cosmological defects or even terminate inflation. We apply the results of a toy scalar model to the case of the Standard Model (SM) Higgs boson, whose quartic evolves to negative values at high scales, and extend previous analyses of Higgs dynamics during inflation utilizing statistical methods to a perturbative and fully gauge-invariant formulation. We show that the dynamics are controlled by the renormalization group-improved quartic coupling $λ(μ)$ evaluated at a scale $μ= H$, such that Higgs fluctuations are enhanced by the instability if $H > Λ_I$. Even if $H > Λ_I$, the instability in the SM Higgs potential does not end inflation; instead the universe slowly sloughs off crunching patches of space that never come to dominate the evolution. As inflation proceeds past 50 $e$-folds, a significant proportion of patches exit inflation in the unstable vacuum, and as much as 1% of the spacetime can rapidly evolve to a defect. Depending on the nature of these defects, however, the resulting universe could still be compatible with ours.

hep-th↗

arXiv2015

Natural Supersymmetry without Light Higgsinos

We present a mechanism that allows a large Higgsino mass without large fine-tuning. The Higgs is a pseudo Nambu-Goldstone boson (PNGB) of the global symmetry breaking pattern $SO(5) \to SO(4)$. Because of the PNGB nature of the light Higgs, the $SO(5)$ invariant Higgsino mass does not directly contribute to the Higgs mass. Large couplings in the Higgs sector that spontaneously breaks $SO(5)$ minimize the tuning, and are also motivated by the requirements of generating a sufficiently large Higgs quartic coupling and of maintaining a natural approximate global $SO(5)$ symmetry. When these conditions are imposed, theories of this type predict heavy Higgsinos. This construction differs from composite Higgs models in that no new particles are introduced to form complete $SO(5)$ multiplets involving the top quark---the stop is the only top partner. Compatibility with Higgs coupling measurements requires cancelations among contributions to the Higgs mass squared parameter at the 10% level. An important implication of this construction is that the compressed region of stop and sbottom searches can still be natural.

hep-ph↗

arXiv2015

Probable or Improbable Universe? Correlating Electroweak Vacuum Instability with the Scale of Inflation

Measurements of the Higgs boson and top quark masses indicate that the Standard Model Higgs potential becomes unstable around $Λ_I \sim 10^{11}$ GeV. This instability is cosmologically relevant since quantum fluctuations during inflation can easily destabilize the electroweak vacuum if the Hubble parameter during inflation is larger than $Λ_I$ (as preferred by the recent BICEP2 measurement). We perform a careful study of the evolution of the Higgs field during inflation, obtaining different results from those currently in the literature. We consider both tunneling via a Coleman-de Luccia or Hawking-Moss instanton, valid when the scale of inflation is below the instability scale, as well as a statistical treatment via the Fokker-Planck equation appropriate in the opposite regime. We show that a better understanding of the post-inflation evolution of the unstable AdS vacuum regions is crucial for determining the eventual fate of the universe. If these AdS regions devour all of space, a universe like ours is indeed extremely unlikely without new physics to stabilize the Higgs potential; however, if these regions crunch, our universe survives, but inflation must last a few e-folds longer to compensate for the lost AdS regions. Lastly, we examine the effects of generic Planck-suppressed corrections to the Higgs potential, which can be sufficient to stabilize the electroweak vacuum during inflation.

hep-ph↗

arXiv2013

Dark Sector Mass Relations from RG Focusing

Dark sector mass relations, such as those which permit near-threshold or near-resonance annihilation in the early universe, could arise due to IR-attractive ratios in renormalization group equations. Achieving a particular ratio requires specific dark matter gauge charges or interactions, leading to predictions about the dark matter properties. Furthermore, additional states with masses comparable to the dark matter mass may be necessary, potentially giving rise to novel phenomenology. We explore this idea in the context of dark matter charged under a new gauged $U(1)_X$ that kinetically mixes with the Standard Model hypercharge.

hep-ph↗

arXiv2013

Exotic Top Partners and Little Higgs

Little Higgs models often give rise to top partners beyond the minimal ones necessary for the cancellation of quadratic divergences. We review how this occurs and discuss the phenomenology of these exotic states. We emphasize the possible importance of new pseudo-Nambu-Goldstone bosons in top partner decays. Indeed, cascade decays of exotic top partners may be the best way to discover these new bosons. We illustrate these points with a new Little Higgs construction based on an SO(10)/SO(5)^2 coset structure, which fills a gap in the model building literature. These observations motivate new search strategies for top partners at the LHC, including for final states with b-jets and a large multiplicity of electroweak bosons.

hep-ph↗

arXiv2013

Top Partner Probes of Extended Higgs Sectors

Natural theories of the weak scale often include fermionic partners of the top quark. If the electroweak symmetry breaking sector contains scalars beyond a single Higgs doublet, then top partners can have sizable branching ratios to these extended Higgs sector states. In fact, top partner decays may provide the most promising discovery mode for such scalars, especially given the large backgrounds to direct and associated production. In this paper, we present a search strategy for top partner decays to a charged Higgs boson and a bottom quark, focusing on the case where the charged Higgs dominantly decays to third-generation quarks to yield a multi-b final state. We also discuss ways to extend this search to exotic neutral scalars decaying to bottom quark pairs.

hep-ph↗

arXiv2012

Vectorlike Fermions and Higgs Couplings

New vectorlike fermions that mix with the third generation can significantly affect the tau and b Yukawa couplings. Consistent with precision electroweak measurements, the width of the Higgs boson to tau tau, b b can be reduced by O(1) with respect to the Standard Model values. In the case of the b quark, a reduced width would result in an enhanced branching ratio for other final states, such as gamma gamma. New leptons can also substantially modify the Higgs boson branching ratio to photons through radiative effects, while new quarks can contribute to g g fusion. The combined effect can be as much as a factor of two on the branching ratio to gamma gamma. The new quarks and leptons could be light, which would allow discovery at the LHC. In the case of significant suppression of h -> tau tau, searches for new leptons decaying to tau-rich final states, perhaps in association with Higgs bosons, are motivated.

hep-ph↗

arXiv2012

Neutrinos from Off-Shell Final States and the Indirect Detection of Dark Matter

We revisit the annihilation of dark matter to neutrinos in the Sun near the WW and tt kinematic thresholds. We investigate the potential importance of annihilation to WW* in a minimal dark matter model in which a Majorana singlet is mixed with a vector-like electroweak doublet, but many results generalize to other models of weakly-interacting dark matter. We re-evaluate the indirect detection constraints on this model and find that, once all annihilation channels are properly taken into account, the most stringent constraints on spin-dependent scattering for dark matter mass 60 GeV < mX < mt are derived from the results of the Super-Kamiokande experiment. Moreover, we establish the model-independent statement that Majorana dark matter whose thermal relic abundance and neutrino signals are both controlled by annihilation via an s-channel Z boson is excluded for 70 GeV < mX < mW. In some models, annihilation to tt* can affect indirect detection, notably by competing with annihilation to gauge boson final states and thereby weakening neutrino signals. However, in the minimal model, this final state is largely negligible, only allowing dark matter with mass a few GeV below the top quark mass to evade exclusion.

hep-ph↗

arXiv2011

Singlet-Doublet Dark Matter

In light of recent data from direct detection experiments and the Large Hadron Collider, we explore models of dark matter in which an SU(2) doublet is mixed with a Standard Model singlet. We impose a thermal history. If the new particles are fermions, this model is already constrained due to null results from XENON100. We comment on remaining regions of parameter space and assess prospects for future discovery. We do the same for the model where the new particles are scalars, which at present is less constrained. Much of the remaining parameter space for both models will be probed by the next generation of direct detection experiments. For the fermion model, DeepCore may also play an important role.

hep-ph↗