Machine learning techniques in particle physics are most powerful when they are trained directly on data, to avoid sensitivity to theoretical uncertainties or an underlying bias on the expected signal. To be able to train on data in searches for new physics, anomaly detection methods are imperative, which can be realised by an autoencoder acting as an unsupervised classifier. The last source of uncertainties affecting the classifier are then experimental uncertainties in the reconstruction of the final-state objects. To mitigate their effect on the classifier and to allow for a realistic assessment of the method, we propose to combine the autoencoder with an adversarial neural network to remove its sensitivity to the smearing of the final-state objects. We quantify its effect and show that one can achieve a robust anomaly detection in resonance-induced tt final states. * If machine learning techniques can be trained on data directly they become independent of theoretical uncertainties. In such circumstances they can outperform theory-based reconstruction approaches, like the matrix element method [44][45][46][47][48], which was recently extended to fully exclusive final states [49][50][51][52][53][54].† The quantum numbers of the decaying resonances are known to have a strong impact on the reconstruction efficiencies of boosted top quarks [61].

Starting from the observation that artificial neural networks are uniquely suited to solving optimisation problems, and most physics problems can be cast as an optimisation task, we introduce a novel way of finding a numerical solution to wide classes of differential equations. We find our approach to be very flexible and stable without relying on trial solutions, and applicable to ordinary, partial and coupled differential equations. We apply our method to the calculation of tunnelling profiles for cosmological phase transitions, which is a problem of relevance for baryogenesis and stochastic gravitational wave spectra. Comparing our solutions with publicly available codes which use numerical methods optimised for the calculation of tunnelling profiles, we find our approach to provide at least as accurate results as these dedicated differential equation solvers, and for some parameter choices even more accurate and reliable solutions. In particular, we compare the neural network approach with two publicly available profile solvers, CosmoTransitions and BubbleProfiler, and give explicit examples where the neural network approach finds the correct solution while dedicated solvers do not. We point out that this approach of using artificial neural networks to solve equations is viable for any problem that can be cast into the form F( x) = 0, and is thus applicable to various other problems in perturbative and non-perturbative quantum field theory.

Publisher's copyright statement:Reprinted with permission from the American Physical Society: Physical Review D 96, 055042 c 2017 by the American Physical Society. Readers may view, browse, and/or download material for temporary copying purposes only, provided these uses are for noncommercial personal purposes. Except as provided by law, this material may not be further reproduced, distributed, transmitted, modied, adapted, performed, displayed, published, or sold in whole or part, without prior written permission from the American Physical Society. Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. Heavy neutrinos, a key prediction of many standard model extensions, remain some of the most searched-for objects at collider experiments. In this context, we revisit the premise that the gluon fusion production mechanism, gg → Z à =h à → Nν l , is phenomenologically irrelevant at the CERN LHC and report the impact of soft gluon corrections to the production cross section. We resum threshold logarithms up to next-to-next-to-next-to-leading logarithmic accuracy (N 3 LL), thus capturing the dominant contributions to the inclusive cross section up to next-to-next-to-leading order (N 2 LO). For m N > 150 GeV and collider energies ffiffi ffi s p ¼ 7-100 TeV, corrections to the Born rates span from þ160% to þ260%. At ffiffi ffi s p ¼ 14 TeV, the resummed channel is roughly equal in size to the widely-believed-to-be-dominant charged-current Drell-Yan process and overtakes it outright at ffiffi ffi s p ≳ 20-25 TeV. Results are independent of the precise nature/mixing of N and hold generically for other low-scale seesaws. Findings are also expected to hold for other exotic leptons and broken axial-vector currents, particularly as the Z à contribution identically reduces to that of a pseudoscalar.

Abstract:We calculate the two-loop contributions from a modified trilinear Higgs selfinteraction, κ λ λ SM vh 3 , to the electroweak oblique parameters S and T . Using the current bounds on S and T from electroweak measurements, we find the 95% C.L. constraint on the modified trilinear coupling to be −14.0 ≤ κ λ ≤ 17.4. The largest effects on S and T arise from two insertions of the modified trilinear coupling that result in T /S −3/2; remarkably, this is nearly parallel to the axis of the tightest experimental constraint in the S-T plane. No contributions to S and T arise from a modified Higgs quartic coupling at two-loop order. These calculations utilized a gauge-invariant parameterization of the trilinear Higgs coupling in terms of higher dimensional operators (H † H) n with n ≥ 3. Interestingly, the bounds on κ λ that we obtain are comparable to constraints from diHiggs production at the LHC as well as recent bounds from single Higgs production at the LHC.

Higgsplosion is the mechanism that leads to exponentially growing decay rates of highly energetic particles into states with very high numbers of relatively soft Higgs bosons. In this paper we study quantum effects in the presence of Higgsplosion. First, we provide a non-perturbative definition of Higgsplosion as a resolved short-distance singularity of quantum propagators at distances shorter than the inverse Higgsplosion energy scale, E * . We then consider quantum effects arising from loops in perturbation theory with these propagators on internal lines. When the loop momenta exceed the Higgsplosion scale E * , the theory dynamics deviates from what is expected in the standard QFT settings without Higgsplosion. The UV divergences are automatically regulated by the Higgsplosion scale, leading to the change of slopes for the running couplings at the RG scales µ > E * . Thus, the theory becomes asymptotically safe. Further, we find that the finite parts are also modified and receive power-suppressed corrections in 1/E 2 * . We use these results to compute a set of precision observables for the Higgsploding Standard Model. These and other precision observables could provide experimental evidence and tests for the existence of Higgsplosion in particle physics. arXiv:1709.08655v3 [hep-ph]