arXiv subjects

Christfried Focke

Publications and source records attributed to Christfried Focke.

7 recordsLinked to original sources

arXiv2019

Fine-tune Bert for DocRED with Two-step Process

Modelling relations between multiple entities has attracted increasing attention recently, and a new dataset called DocRED has been collected in order to accelerate the research on the document-level relation extraction. Current baselines for this task uses BiLSTM to encode the whole document and are trained from scratch. We argue that such simple baselines are not strong enough to model to complex interaction between entities. In this paper, we further apply a pre-trained language model (BERT) to provide a stronger baseline for this task. We also find that solving this task in phases can further improve the performance. The first step is to predict whether or not two entities have a relation, the second step is to predict the specific relation.

cs.CL↗

arXiv2016

$W$-boson production in association with a jet at next-to-next-to-leading order in perturbative QCD

We present the complete calculation of $W$-boson production in association with a jet in hadronic collisions through next-to-next-to-leading order in perturbative QCD. To cancel infrared divergences we introduce a new subtraction method that exploits the fact that the $N$-jettiness event-shape variable fully captures the singularity structure of QCD amplitudes with final-state partons. This method holds for processes with an arbitrary number of jets, and is easily implemented into existing frameworks for higher-order calculations. We present initial phenomenological results for $W$+jet production at the LHC. The NNLO corrections are small and lead to a significantly reduced theoretical error, opening the door to precision measurements in the $W$+jet channel at the LHC.

hep-ph↗

arXiv2016

Z-boson production in association with a jet at next-to-next-to-leading order in perturbative QCD

We present the first complete calculation of Z-boson production in association with a jet in hadronic collisions through next-to-next-to-leading order in perturbative QCD. Our computation uses the recently-proposed N-jettiness subtraction scheme to regulate the infrared divergences that appear in the real-emission contributions. We present phenomenological results for 13 TeV proton-proton collisions with fully realistic fiducial cuts on the final-state particles. The remaining theoretical uncertainties after the inclusion of our calculations are at the percent-level, making the Z+jet channel ready for precision studies at the LHC Run II.

hep-ph↗

arXiv2016

Color singlet production at NNLO in MCFM

We present the implementation of several color-singlet final-state processes at Next-to-Next-to Leading Order (NNLO) accuracy in QCD to the publicly available parton-level Monte Carlo program MCFM. Specifically we discuss the processes $pp\rightarrow H$, $pp\rightarrow Z$, $pp\rightarrow W$, $pp\rightarrow HZ$, $pp\rightarrow HW$ and $pp\rightarrowγγ$. Decays of the unstable bosons are fully included, resulting in a flexible fully differential Monte Carlo code. The NNLO corrections have been calculated using the non-local $N$-jettiness subtraction approach. Special attention is given to the numerical aspects of running MCFM for these processes at this order. We pay particular attention to the systematic uncertainties due to the power corrections induced by the $N$-jettiness regularization scheme and the evaluation time needed to run the hybrid openMP/MPI version of MCFM at NNLO on multi-processor systems.

hep-ph↗

arXiv2015

Jet vetoes versus giant K-factors in the exclusive Z+1-jet cross section

We perform a detailed study of the exclusive Z+1-jet cross section at the 13 TeV LHC, motivated by the importance of similar exclusive cross sections in understanding the production of the Higgs boson in the W^+W^- final state. We point out a feature of the ATLAS analysis that has significant impact on the theoretical predictions: the jet-isolation criterion implemented by ATLAS effectively allows dijet events where an energetic jet is collinear to a final-state lepton. This process contains a giant $K$-factor arising from the collinear emission of a Z-boson from the dijet configuration. This overwhelms the effect of the jet-veto logarithms, making it difficult to test their resummation in this process. We provide numerical results that demonstrate the interplay between the jet-veto logarithms and the giant K-factor in the theoretical prediction. We study several observables, including the transverse momentum distributions of the leading jet and the Z-boson, in the exclusive Z+1-jet process, and discuss their sensitivity to both the giant K-factor and the jet-veto logarithms. We suggest a modified isolation criterion that removes the giant K-factor and allows for a direct test of the jet-veto resummation framework in the exclusive Z+1-jet process.

hep-ph↗

arXiv2015

Higgs boson production in association with a jet at NNLO using jettiness subtraction

We use the recently proposed jettiness-subtraction scheme to provide the complete calculation of Higgs boson production in association with a jet in hadronic collisions through next-to-next-to-leading order in perturbative QCD. This method exploits the observation that the $N$-jettiness event-shape variable completely describes the singularity structure of QCD when final-state colored particles are present. Our results are in agreement with a recent computation of the $gg$ and $qg$ partonic initial states based on sector-improved residue subtraction. We present phenomenological results for both fiducial cross sections and distributions at the LHC.

hep-ph↗

arXiv2014

Jet vetoes for Higgs production at future hadron colliders

We study Higgs boson production in exclusive jet bins at possible future 33 and 100 TeV proton-proton colliders. We compare the cross sections obtained using fixed-order perturbation theory with those obtained by also resuming large logarithms induced by the jet-binning in the gluon-fusion and associated production channels. The central values obtained by the best-available fixed-order predictions differ by $10-20\%$ from those obtained after including resummation over the majority of phase-space regions considered. Additionally, including the resummation dramatically reduces the residual scale variation in these regions, often by a factor of two or more. We further show that in several new kinematic regimes that can be explored at these high-energy machines, the inclusion of resummation improvement is mandatory.

hep-ph↗