arXiv subjects

Yuta Mukobara

Publications and source records attributed to Yuta Mukobara.

3 recordsLinked to original sources

arXiv2026

Multiquark clustering in neutron-star matter from color-spin molecular dynamics

We study the equation of state of neutron-star matter with color-spin molecular dynamics. The calculation includes the internal color and spin degrees of freedom and their time evolution. The matter composition, including strangeness under $β$ equilibrium, is determined by energy minimization. We find two main trends. First, within the present color-spin molecular dynamics framework and under the adopted clustering criterion along the stable neutron-star branch, isolated quarklike configurations do not appear; instead, color-magnetic interactions favor the self-consistent formation of multiquark clusters. Within the same criterion, the cluster-size distribution is concentrated at quark numbers that are multiples of three, corresponding to integer baryon numbers. Second, relative to the conventional no-$K^*$ baseline, the interaction between strange and light quarks has a strong impact on neutron-star radii. This suggests that future radius measurements, together with phenomenological information on the strangeness-onset density, may help constrain flavor-sector interactions involving strangeness.

astro-ph.HE↗

arXiv2026

A physics-embedded Bayesian neural network for predicting the energy dependence of fission product yields with fine structures

We present a physics-embedded Bayesian neural network (PE-BNN) framework that integrates fission product yields (FPYs) with prior nuclear physics knowledge to predict energy-dependent FPY data with fine structure. By incorporating an energy-independent phenomenological shell factor as a single input feature, the PE-BNN captures both fine structures and global energy trends. The combination of this physics-informed input with hyperparameter optimization via the Watanabe-Akaike Information Criterion (WAIC) significantly enhances predictive performance. Our results demonstrate that the PE-BNN framework is well-suited for target observables with systematic features that can be embedded as model inputs, achieving close agreement with known shell effects and prompt neutron multiplicities.

nucl-th↗

arXiv2024

Rethinking Loss Functions for Fact Verification

We explore loss functions for fact verification in the FEVER shared task. While the cross-entropy loss is a standard objective for training verdict predictors, it fails to capture the heterogeneity among the FEVER verdict classes. In this paper, we develop two task-specific objectives tailored to FEVER. Experimental results confirm that the proposed objective functions outperform the standard cross-entropy. Performance is further improved when these objectives are combined with simple class weighting, which effectively overcomes the imbalance in the training data. The souce code is available at https://github.com/yuta-mukobara/RLF-KGAT

cs.CL↗