Abstract:We study the effects produced by D-brane instantons on the holomorphic quantities of a D-brane gauge theory at an orbifold singularity. These effects are not limited to reproducing the well known contributions of the gauge theory instantons but also generate extra terms in the superpotential or the prepotential. On these brane instantons there are some neutral fermionic zero-modes in addition to the ones expected from broken supertranslations. They are crucial in correctly reproducing effects which are dual to gauge theory instantons, but they may make some other interesting contributions vanish. We analyze how orientifold projections can remove these zero-modes and thus allow for new superpotential terms. These terms contribute to the dynamics of the effective gauge theory, for instance in the stabilization of runaway directions.
In this paper we show that it is possible to derive non-perturbative superpotential terms from a stringy instanton without introducing orientifold planes. The instanton is realized by a Euclidean D brane wrapping a non-trivial cycle upon which we also wrap a single space-filling D brane. The standard problem of unwanted neutral fermionic zero modes is evaded by the appearance of couplings to charged bosonic zero modes in the instanton moduli action. Since the Euclidean D brane wraps a cycle which is not associated to any low energy gauge dynamics, it can not be interpreted as an ordinary gauge instanton, but rather as a stringy one. By considering such a brane configuration at an orbifold singularity, we can explicitly evaluate the instanton moduli space integral and find a holomorphic superpotential term with the structure of a baryonic mass term.
Due to the difficulties of finding superconformal Lagrangian theories for multiple M2-branes, we will in this paper instead focus on the field equations. By relaxing the requirement of a Lagrangian formulation we can explore the possibility of having structure constants f ABC D satisfying the fundamental identity but which are not totally antisymmetric. We exemplify this discussion by making use of an explicit choice of a non-antisymmetric f ABC D constructed from the Lie algebra structure constants f ab c of an arbitrary gauge group. Although this choice of f ABC D does not admit an obvious Lagrangian description, it does reproduce the correct SYM theory for a stack of N D2-branes to leading order in g −1 YM upon reduction and, moreover, it sheds new light on the centre of mass coordinates for multiple M2-branes.
We interpret the di-photon excess recently reported by the ATLAS and CMS collaborations as a new resonance arising from the sgoldstino scalar, which is the superpartner of the Goldstone mode of spontaneous supersymmetry breaking, the goldstino. The sgoldstino is produced at the LHC via gluon fusion and decays to photons, with interaction strengths proportional to the corresponding gaugino masses over the supersymmetry breaking scale. Fitting the excess, while evading bounds from searches in the di-jet, Zγ, ZZ and W W final states, selects the supersymmetry breaking scale to be a few TeV, and particular ranges for the gaugino masses. The two real scalars, corresponding to the CP-even and CP-odd parts of the complex sgoldstino, both have narrow widths, but their masses can be split of the order 10-30 GeV by electroweak mixing corrections, which could account for the preference of a wider resonance width in the current low-statistics data. In the parameter space under consideration, tree-level F -term contributions to the Higgs mass arise, in addition to the standard D-term contribution proportional to the Z-boson mass, which can significantly enhance the tree level Higgs mass.
We introduce an automated tool for deploying ultra low-latency, low-power deep neural networks with convolutional layers on field-programmable gate arrays (FPGAs). By extending the hls4ml library, we demonstrate an inference latency of 5 µs using convolutional architectures, targeting microsecond latency applications like those at the CERN Large Hadron Collider. Considering benchmark models trained on the Street View House Numbers Dataset, we demonstrate various methods for model compression in order to fit the computational constraints of a typical FPGA device used in trigger and data acquisition systems of particle detectors. In particular, we discuss pruning and quantization-aware training, and demonstrate how resource utilization can be significantly reduced with little to no loss in model accuracy. We show that the FPGA critical resource consumption can be reduced by 97% with zero loss in model accuracy, and by 99% when tolerating a 6% accuracy degradation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.