arXiv subjects

Zilin Yuan

Publications and source records attributed to Zilin Yuan.

7 recordsLinked to original sources

arXiv2026

Exploring the chiral magnetic effect in Au+Au collisions at $\sqrt{s_{NN}}=7.7-200$ GeV through Chiral Anomaly Transport

High-energy heavy-ion collisions have the potential to create local domains of chirality-imbalanced quarks, reflecting the topological characteristics of quantum chromodynamics. This phenomenon can potentially induce local $\mathcal{P}$ and $\mathcal{CP}$ violations in the quark-gluon plasma. The Chiral Magnetic Effect (CME) predicts an electric charge separation along the intense magnetic field generated during these collisions, which is typically investigated through charge-dependent azimuthal correlations ($Δγ$). In this work, we investigate the CME in Au+Au collisions at $\sqrt{s_{NN}} = 7.7 - 200$ GeV using a multiphase transport (AMPT) model equipped with a Chiral Anomaly Transport (CAT) module. we employ two independent methods: direct subtraction of the correlator $\langle N_{part}Δγ\rangle$ between simulations with zero and finite chiral chemical potential $μ_5$, and the event-shape-selection (ESS) approach. Our results reveal a significant CME signal within the energy range of 11.5-27 GeV and the centrality range of $20-50\%$, where the AMPT model aligns well with STAR experimental data. Furthermore, the CME fractions extracted by both methods are consistent within uncertainties across these energies. However, the CME signal disappears at both 7.7 and 200 GeV. These findings underscore that the observability of the CME critically depends on both the dynamic evolution of the magnetic field and the chemical freeze-out time of the partonic phase, which vary significantly with collision energy.

hep-ph↗

arXiv2026

Affordance2Action: Task-Conditioned Scene-level Affordance Grounding for Real-Time Manipulation

Task-conditioned manipulation requires grounding instructions to task-relevant functional parts rather than object categories. This setting is scene-dependent and often one-to-many in cluttered scenes: the same object may afford different interactions across tasks, while a single task may correspond to either one functional region or multiple valid functional regions, depending on the scene layout. Existing affordance datasets and benchmarks remain misaligned with this setting, as they typically focus on grasping or object-level affordances, rely on synthetic scenes, or assume a single instruction-region correspondence. We present Affordance2Action (A2A), a benchmark-centered learning framework for scene-level, task-conditioned part affordance grounding. At its core is A2A-Bench, a manipulation-oriented benchmark that covers both single-region and multi-region instruction correspondences in everyday scenes, with the latter highlighting the ambiguity and diversity of affordance grounding in realistic multi-object environments. To construct it at scale, we build A2A-AffordGen, an agent-assisted annotation pipeline that combines language-model filtering, interactive part segmentation, instance-level mask-out refinement, task-reasoning instruction generation, and human verification. A2A-Bench's supervision further supports diverse downstream applications, with real-time affordance grounding and affordance-conditioned manipulation policies as two representative examples. Experiments show that A2A exposes substantial gaps in generic segmentation, VLM-based grounding, and affordance distillation baselines, while improving task-level localization and providing useful spatial priors for downstream manipulation. All datasets and code will be publicly released to promote open research.

cs.RO↗

arXiv2024

Exploring the chiral magnetic effect in isobar collisions through Chiral Anomaly Transport

We investigate the signal of the chiral magnetic effect (CME) in Au+Au collisions and isobar collisions of $_{44}^{96}\text{Ru}+\rm{} _{44}^{96}Ru$ and $_{40}^{96}\text{Zr}+\rm{}_{40}^{96}Zr$ in the newly developed chiral anomaly transport (CAT) module based on the state-of-the-art model a multiphase transport (AMPT). Our numerical simulation results for the ratio charge correlation $Δγ$ in Ru+Ru and Zr+Zr collisions are close to the latest experimental data. The simulation shows that the CME signal is larger in Ru+Ru collisions than that in Zr+Zr collisions, while the background is smaller, and the upper limit of the CME signal is $15\%$ in the isobar collisions.

hep-ph↗

arXiv2024

Dynamical induced quark spin polarization by magnetic field at the early stage of heavy-ion collisions

We present a comprehensive analysis of the dynamic process of quark spin polarization induced by magnetic fields at the pre-thermal stage in heavy-ion collisions by using the recently developed theoretical tool of chiral kinetic theory. Our findings demonstrate that the spin polarization of quarks is highly sensitive to the interactions between quarks. These interactions can delay the decay of early spin polarization vector while accelerating the decay of later spin polarization vector. Specifically, our simulations show the detailed process of how magnetic fields polarize quarks within the fireball and reveal that quark interactions lead to an acceleration effect on the average spin. Notably, the fireball of quark-gluon plasma (QGP) in its early stages exhibits an incomplete electromagnetic response effect, which differs from the response predicted by Lenz's law. This discrepancy arises from quantum corrections involving the interactions between quark spin and electromagnetic fields.

nucl-th↗

arXiv2023

An Anchor Learning Approach for Citation Field Learning

Citation field learning is to segment a citation string into fields of interest such as author, title, and venue. Extracting such fields from citations is crucial for citation indexing, researcher profile analysis, etc. User-generated resources like academic homepages and Curriculum Vitae, provide rich citation field information. However, extracting fields from these resources is challenging due to inconsistent citation styles, incomplete sentence syntax, and insufficient training data. To address these challenges, we propose a novel algorithm, CIFAL (citation field learning by anchor learning), to boost the citation field learning performance. CIFAL leverages the anchor learning, which is model-agnostic for any Pre-trained Language Model, to help capture citation patterns from the data of different citation styles. The experiments demonstrate that CIFAL outperforms state-of-the-art methods in citation field learning, achieving a 2.68% improvement in field-level F1-scores. Extensive analysis of the results further confirms the effectiveness of CIFAL quantitatively and qualitatively.

cs.CL↗

arXiv2023

Evolution of topological charge through chiral anomaly transport

Built upon the state-of-the-art model a multiphase transport (AMPT), we develop a new module of chiral anomaly transport (CAT), which can trace the evolution of the initial topological charge of gauge field created through sphaleron transition at finite temperature and external magnetic field in heavy ion collisions. The eventual experimental signals of chiral magnetic effect(CME) can be measured. The CAT explicitly shows the generation and evolution of the charge separation, and the signals of CME through the CAT are quantitatively in agreement with the experimental measurements in Au+Au collision at $\sqrt{s}=200 {\rm GeV}$, and the centrality dependence of the CME fraction follows that of the fireball temperature.

hep-ph↗

arXiv2022

A Curriculum Learning Approach for Multi-domain Text Classification Using Keyword weight Ranking

Text classification is a very classic NLP task, but it has two prominent shortcomings: On the one hand, text classification is deeply domain-dependent. That is, a classifier trained on the corpus of one domain may not perform so well in another domain. On the other hand, text classification models require a lot of annotated data for training. However, for some domains, there may not exist enough annotated data. Therefore, it is valuable to investigate how to efficiently utilize text data from different domains to improve the performance of models in various domains. Some multi-domain text classification models are trained by adversarial training to extract shared features among all domains and the specific features of each domain. We noted that the distinctness of the domain-specific features is different, so in this paper, we propose to use a curriculum learning strategy based on keyword weight ranking to improve the performance of multi-domain text classification models. The experimental results on the Amazon review and FDU-MTL datasets show that our curriculum learning strategy effectively improves the performance of multi-domain text classification models based on adversarial learning and outperforms state-of-the-art methods.

cs.CL↗