핵·입자물리 분야 슈퍼컴퓨팅 활용 거대문제 연구를 발굴하기 위하여 고에너지물리 거대컴퓨팅 및 머신러닝 관련 연구자들이 참석하여 포럼을 개최한다. 또한, 요사이 핫 이슈가 되고 있는 암흑물질에 관하여 현재 진행사항과 향후 전망에 대하여 논의한다.
양자 메트릭 학습(Quantum Metric Learning)은 자기지도학습의 하나로 분류될 수 있다. 라벨링 되지 않은 정보를 큐빗에 반복해서 입력함으로써, 샘플 사이의 거리를 벌리는 방식으로 분류하고 학습할 수 있기 때문이다. 이 방법을 사용하여 LHC에서 투 힉스 더블렛 모델의 CP-odd scalar A가 Z보존과 힉스를 통해 젯으로 붕괴하는 프로세스의 시그널을 구분 할 수 있다. 유사하게, 고전적 대조학습 중에서 메트릭 학습 방법을 사용하는 것 중 하나인 SimCLR(A simple framework for contrastive learning of visual representation)과 비교될 수 있는데, 양자 메트릭 학습 방법과 성능을 비교해 본다.
U(1)_(Lμ-Lτ) ≡ U(1)_X model is anomaly free within the Standard Model (SM) fermion content, and can accommodate the muon (g−2) data for M_Z' ~ O(10-100) MeV and g_X ~ (4-8)×10^(-4). WIMP type thermal dark matter (DM) can be also introduced for M_Z'~2M_DM, if DM pair annihilations into the SM particles occur only through the s-channel Z' exchange. In this work, we show that this tight correlation between M_Z' and M_DM can be completely evaded both for scalar and fermionic DM, if we include the contributions from dark Higgs boson (H_1). Dark Higgs boson plays a crucial role in DM phenomenology, not only for generation of dark photon mass, but also opening new channels for DM pair annihilations into the final states involving dark Higgs boson, such as dark Higgs pair as well as Z'Z' through dark Higgs exchange in the s-channel, and co-annihilation into Z'H_1 in case of inelastic DM. Thus dark Higgs boson will dissect the strong correlation M_Z'~2M_DM, and much wider mass range is allowed for U(1)_X-charged complex scalar and Dirac fermion DM, still explaining the muon (g-2). We consider both generic U(1)_X breaking as well as U(1)_X → Z_2 (and also into Z_3 for scalar DM case).
We will talk about our efforts for numerically estimating the QCD axion dark matter abundance from the topological cosmic string network. Using various state-of-art lattice simulations including the Adaptive Mesh Refinement (AMR), we establish the recently discovered scaling solution and to derive the power law of the axion power spectrum. We will describe the up-to-date status of the cosmic string simulation and existing issues, and finally will present some of our results.
In this talk, I want to discuss recent lattice efforts on the non-perturbative studies of strongly coupled gauge theories other than QCD towards phenomenological model buildings for physics beyond the standard model. Particular emphases are on the light scalar state, identified as a dilaton, in many-flavor QCD and on the baryonic states composed of fermion constituents in the two different representations, called chimera baryons.
Nuclear Lattice Effective Filed Theory is an ab-initio approach for the many-body quantum system. To apply the Nuclear Lattice Effective Field Theory with higher order interactions in Chiral Effective Theory to neutron rich nuclei, new method called Wave Function Matching is developed. The WFM method is applied for the binding energy of light nuclei, medium-mass nuclei, neutron matter and nuclear matter and charge radius of light and medium-mass nuclei. We use interactions at Next-to-nextto-next-to-leading order in chiral effective theory and find good agreement with empirical data.
Ultra-light (fuzzy) dark matter is a promising alternative to the cold dark matter (CDM) model, as it has the potential to solve some of the small-scale issues of CDM. However, the wave nature of fuzzy dark matter makes it challenging to simulate cosmic structure formation, especially on scales larger than galaxies. In this talk, we review various public packages that can be used to simulate fuzzy dark matter, such as Axionyx, Enzo, and GRChombo for large scale structures. As an example, we numerically show that fuzzy dark matter can solve the final parsec problem of binary supermassive black holes.
Belle II is an experiment of colliding energy-asymmetric e+ e- beams using SuperKEKB collider and Belle II detector. While the focus of Belle II experiment is on flavor physics and CP violations, e.g. in B, D mesons and $\tau$ leptons, Belle II is also poised to searching for physics signals of dark sector in the light-mass regime of around a few GeV or less. In this talk, we present recent dark-sector-related studies from Belle II, in which both continuum processes as well as B decay processes are studied.
The search for dark matter with the CMS detector has been performed at CERN LHC. If dark matter particles are produced at the LHC, they would escape the detectors without being directly detected. This presence can be inferred by measuring an imbalance in the transverse energy after the proton-proton collision. Therefore, one of the methods employed at the CMS experiment to search for dark matter is by studying missing transverse energy (MET). We present an improvement of the MET resolution using transformer architecture for regression based on the self-attention mechanism.
우주를 구성하는 물질의 대부분을 차지하고 있는 암흑물질(~25 %) 및 암흑에너지(~70 %)를 탐색하는 것은 21세기 물리학 연구의 주요 과제 중의 하나이다. 현재 지하 실험을 통하여 암흑물질을 직접 검출을 시도하는 연구가 많이 수행되고 있으며, Large Hadron Collider(LHC)과 같은 가속기 실험에서는 고에너지 빔의 충돌에 의하여 생성될 수 있는 암흑물질을 탐색하는 연구가 진행되고 있다. LHC는 2010년 이후 RUN1과 RUN2를 거쳐 2022년부터는 13.6 TeV의 에너지에서 RUN3를 가동하고 있다. LHC의 CMS 실험 그룹이 획득한 데이터를 분석하여 수행한 암흑물질 탐색의 연구 결과를 소개하고 향후 전망에 대하여 발표한다.
SND@LHC is a compact and stand-alone experiment to perform measurements with neutrinos produced at the LHC in a hitherto unexplored pseudo-rapidity region of 7.2 < 𝜂 < 8.6, complementary to all the other experiments at the LHC. The experiment is located 480 m downstream of IP1 in the unused TI18 tunnel. The detector is composed of a hybrid system based on an 800 kg target mass of tungsten plates, interleaved with emulsion and electronic trackers, followed downstream by a calorimeter and a muon system. The configuration allows efficiently distinguishing between all three neutrino flavours, opening a unique opportunity to probe physics of heavy flavour production at the LHC in the region that is not accessible to ATLAS, CMS and LHCb. This region is of particular interest also for future circular colliders and for predictions of very high-energy atmospheric neutrinos. The detector concept is also well suited to searching for Feebly Interacting Particles via signatures of scattering in the detector target. The first phase aims at operating the detector throughout LHC Run 3 to collect a total of 290 fb−1. The experiment was recently installed in the TI18 tunnel at CERN and has seen its first data. A new era of collider neutrino physics is just starting.
We report the optimized experimental requirements to determine neutrino mass hierarchy using electron antineutrinos ( ν¯e ) generated in a nuclear reactor. The features of the neutrino mass hierarchy can be extracted from the |Δm312| and |Δm322| oscillations by applying the Fourier sine and cosine transforms to the L/E spectrum. To determine the neutrino mass hierarchy above 90% probability, the requirements on the energy resolution as a function of the baseline are studied at sin22θ13=0.1 . Suppose the energy resolution of the neutrino detector is less than 0.04/Eν and the determination probability obtained from Bayes' theorem is above 90%. In that case, the detector needs to be located around 48–53 km from the reactor(s) to measure the energy spectrum of ν¯e . These results will help set up an experiment to determine the neutrino mass hierarchy, a critical problem in neutrino physics.