K. M. Graczyk

K. M. Graczyk

Portfolio

Single pion production (SPP) in neutrino interactions with nucleons

Single pion production (SPP) induced by neutrino-nucleon scattering is one of the processes used to measure the neutrino oscillation parameters. We have proposed several improvements in the theoretical description of the SPP. Inpapers Phys. Rev. D77, 053001 and Phys. Rev. D77, 053003, a new scheme for modeling resonance form factors in the Rein-Sehgal model and an algorithm for implementing the lepton mass effects were proposed. Several experimental Monte Carlo generators, including NEUT implemented our results. In Phys. Rev. D 80, 093001, we studied the ANL and BNL data for SPP. For the first time, it was shown that both data sets are consistent, contrary to what was claimed before in previous analyses. In Phys.Rev. D80, 093001, and Phys. Rev. D90, 093001, we obtained new parametrizations for the weak nucleon-Delta excitation transition matrix element.

Electromagnetic and weak structure of the nucleon studied within Bayesian neural network methods

We adapted and developed the Bayesian neural network (BNN) approach to study nucleon’s electroweak internal structure. This method allows reducing the dependence of analysis results on model assumptions. In paper JHEP 09, 053 (2010), we obtained a new parametrization of the proton and neutron’s electromagnetic form factors, and in Phys.Rev.C 99 (2019), 025204 an analogous study was done for the axial form factor. The latter paper showed that neutrino-deuteron scattering data do not contain enough information to study non-dipole corrections. In papers Phys. Rev. C84, 034314; J. Phys. G42, 034019, we studied the two-photon exchange effect. In paper Phys. Rev. C88, 065205 results of the BNN approach were compared to the quantum field theoretical computations. Finally, in Phys. Rev. C90, 054334, we used the Bayesian objective algorithm to calculate the charged proton radius. In addition to the papers’ results, BNN C++ library (written from scratch by K. Graczyk and C. Juszczak) resulted from this study.

Deep learning in porous media

In paper Sci. Rep 10, 21488 (2020), together with Maciej Matyka we proved the deep learning systems’ ability to encode information about porous media fluid dynamics. We adapted the convolutional neural networks (CNN) to relate initial configurations of obstacles, represented by a picture, with three fundamental quantities in the porous media: porosity, permeability, and tortuosity. We showed that our CNN reproduces the values of porosity, permeability, and tortuosity with reasonable accuracy.

Superalgebras for Supergravity


The project aims to find the superalgebra structure that describes intrinsic symmetry. We start from the Poincare and Anti-de Sitter structures. We extend the primary structures by exploiting the so-called resonant construction by adding additional symmetry generators. We analyze millions of superalgebra candidates to find the generator configurations that obey the Poincare or AdS-like super structures pattern. The successful superalgebra structures must satisfy the Jacobi identities.

  • In Eur.Phys.J.C 82 (2022) 3, 254, we obtained symmetry algebras, evaluated within an efficient pattern-based computational method implemented in Wolfram Mathematica. These supersymmetric extensions of algebras, going beyond the Poincaré and Anti-de Sitter ones, contain additional bosonic generators Z_{ab}Zab​ (Lorentz-like), and U_aUa​ (translational-like) added to the standard Lorentz generator J_{ab}Jab​ and translation generator P_{a}Pa​. Our analysis includes all cases up to two fermionic supercharges, Q_{lpha }Qα​ and Y_{lpha }Yα​. The delivered plethora of superalgebras includes a few past results and offers a vastness of new examples. The list of the cases is complete and contains all superalgebras up to two of Lorentz-like, translation-like, and supercharge-like generators (JP+Q)+(ZU+Y)=JPZU+QY(JP+Q)+(ZU+Y)=JPZU+QY. In the latter class, among 667 founded superalgebras, 264 are suitable for direct supergravity construction. For each of them, one can construct a unique supergravity model defined by the Lagrangian. As an example, we consider one of the algebra configurations and provide its Lagrangian realization.
  • In Phys.Lett.B 833 (2022) 137366, we present new superalgebra for N=2D=3,4 supergravity theory endowed with the U(1) generator. The superalgebra is rooted in the so-called Soroka-Soroka algebra and is spanned by the Lorentz Jab and Lorentz-like Zab, translation Pa, T generators, and two supercharges QαI. It is the only possible realization for a given generator content. We construct a corresponding 3D Chern-Simons supergravity realization of the superalgebra and discuss its relevance.

    The project has been done with Remigiusz Durka.

Polarization asymmetries in neutrino scattering off nucleon


The main goal of the project was to study the information content of the polarization asymmetries in:


The project was done with Beata Kowal.

Target spin asymmetry and double and triple spin asymmetries have been studied for the first time for the QE scattering. We showed that double spin asymmetries and triple spin asymmetry contain information about the axial content of the nucleon and are perfect observables for measuring the axial form factor of the nucleon.

One of the problems in modeling the SPP in neutrino-nucleon interactions is a proper description of the resonant (RES) and non-resonant (NR) contribution to the neutrino scattering cross-sections. We showed that the target spin asymmetries and recoil nucleon polarization contain nontrivial information about the interference of resonant and non-resonant amplitudes. In particular, we showed that measurement of the target spin asymmetries might carry out the information about the relative phase between RES and NR amplitudes as well as the NR amplitudes. We also studied the sensitivity of spin asymmetries on the parameters of the N-Delta(1232) weak transition model.