Search

High thermoelectric properties are associated with the phonon-glass electron-crystal paradigm. Conventional wisdom suggests that the optimal bandgap of semiconductor to achieve the largest power factor should be between 6 and 10 kbT. To address challenges related to the bipolar effect and temperature limitations, we present findings on Zintl-type TlAgI2, which demonstrates an exceptionally low lattice thermal conductivity of 0.3 W m-1 K-1 at 300 K. The achieved figure of merit (ZT) for TlAgI2, featuring a 1.55 eV bandgap, reaches a value of 2.20 for p-type semiconductor. This remarkable ZT is attributed to the existence of extended antibonding states Ag-I in the valence band. Furthermore, the bonding hierarchy, influencing phonon anharmonicity, and coordination bonds, facilitating electron transfer between the ligand and the central metal ion, significantly contribute to electronic transport. This finding serves as a promising avenue for the development of high ZT materials with wide bandgaps at elevated temperatures.

Measuring the value of individual samples is critical for many data-driven tasks, e.g., the training of a deep learning model. Recent literature witnesses the substantial efforts in developing data valuation methods. The primary data valuation methodology is based on the Shapley value from game theory, and various methods are proposed along this path. {Even though Shapley value-based valuation has solid theoretical basis, it is entirely an experiment-based approach and no valuation model has been constructed so far.} In addition, current data valuation methods ignore the interpretability of the output values, despite an interptable data valuation method is of great helpful for applications such as data pricing. This study aims to answer an important question: is data valuation learnable and interpretable? A learned valuation model have several desirable merits such as fixed number of parameters and knowledge reusability. An intrepretable data valuation model can explain why a sample is valuable or invaluable. To this end, two new data value modeling frameworks are proposed, in which a multi-layer perception~(MLP) and a new regression tree are utilized as specific base models for model training and interpretability, respectively. Extensive experiments are conducted on benchmark datasets. {The experimental results provide a positive answer for the question.} Our study opens up a new technical path for the assessing of data values. Large data valuation models can be built across many different data-driven tasks, which can promote the widespread application of data valuation.

In this research, we conduct a global QCD analysis of fragmentation functions (FFs) for neutral pions ($\pi^0$), neutral kaons ($K_S^0$), and eta mesons ($\eta$), utilizing world data of single inclusive hadron production in $e^+e^-$ annihilation involving the most recent BESIII data with low collision energy, to test the operational region of QCD collinear factorization for single inclusive hadron production. We found that the QCD-based analysis at next-to-next-to leading order in perturbative QCD with parameterized higher-twist effects can explain both existing high-energy world data and the BESIII new measurements, while the latter cannot be explained with existing FFs extracted with high-energy data. To investigate the higher-twist contributions to this discrepancy, a direct functional approach is employed, providing testing framework for characterizing the experimental results over a wide range of energy scales, from low to high, thus extending the classical theoretical models to the BESIII domain., Comment: 14 pages, 9 figures

Features, logits, and labels are the three primary data when a sample passes through a deep neural network. Feature perturbation and label perturbation receive increasing attention in recent years. They have been proven to be useful in various deep learning approaches. For example, (adversarial) feature perturbation can improve the robustness or even generalization capability of learned models. However, limited studies have explicitly explored for the perturbation of logit vectors. This work discusses several existing methods related to class-level logit perturbation. A unified viewpoint between positive/negative data augmentation and loss variations incurred by logit perturbation is established. A theoretical analysis is provided to illuminate why class-level logit perturbation is useful. Accordingly, new methodologies are proposed to explicitly learn to perturb logits for both single-label and multi-label classification tasks. Extensive experiments on benchmark image classification data sets and their long-tail versions indicated the competitive performance of our learning method. As it only perturbs on logit, it can be used as a plug-in to fuse with any existing classification algorithms. All the codes are available at https://github.com/limengyang1992/lpl.

Measuring the pionic structure function is of high interests as it provides a new area for understanding the strong interaction among quarks and testing the QCD predictions. To this purpose, we investigate the feasibility and the expected impacts of a possible experiment on EicC. We show the simulation results on the statistical precision of an EicC measurement, based on the model of leading neutron tagged DIS process and the parton distribution functions of the pion from JAM18 global analysis. The simulation shows that at EicC, the kinematics cover $x_{\pi}$ range from 0.01 to 1, and $Q^2$ range from 1 GeV$^2$ to 50 GeV$^2$, within the acceptable statistical uncertainty. Assuming an integrated luminosity of 50 fb$^{-1}$, in the low-$Q^{2}$ region ($<10$ GeV$^2$), the MC data show that the suggested measurement in the whole $x_{\rm\pi}$ range reaches very high precision ($<3$\%). To perform such an experiment, only the addition of a far-forward neutron calorimeter is needed., Comment: 9 pages, 12 figures