arXiv subjects

Meng Gao

Publications and source records attributed to Meng Gao.

45 records · Page 3Linked to original sources

arXiv2021

Effects of minor alloying on the mechanical properties of Al based metallic glasses

Minor alloying is widely used to control mechanical properties of metallic glasses (MGs). The present understanding of how a small amount of alloying element changes strength is that the additions lead to more efficient packing of atoms and increased local topological order, which then increases the barrier for shear transformations and the resistance to plastic deformation. Here, we discover that minor alloying can improve the strength of MGs by increasing the chemical bond strength alone and show that this strengthening is distinct from changes in topological order. The results were obtained using Al-Sm based MGs minor alloyed with transition metals (TMs). The addition of TMs led to an increase in the hardness of the MGs which, however, could not be explained based on changes in the topological ordering in the structure. Instead we found that it was the strong bonding between TM and Al atoms which led to a higher resistance to shear transformation that resulted in higher strength and hardness, while the topology around the TM atoms had no influence on their mechanical response. This finding demonstrates that the effects of topology and chemistry on mechanical properties of MGs are independent of each other and that they should be understood as separate, sometimes competing mechanisms of strengthening. This understanding lays a foundation for design of MGs with improved mechanical properties.

cond-mat.mtrl-sci↗

arXiv2021

Exploring Semantic Relationships for Unpaired Image Captioning

Recently, image captioning has aroused great interest in both academic and industrial worlds. Most existing systems are built upon large-scale datasets consisting of image-sentence pairs, which, however, are time-consuming to construct. In addition, even for the most advanced image captioning systems, it is still difficult to realize deep image understanding. In this work, we achieve unpaired image captioning by bridging the vision and the language domains with high-level semantic information. The motivation stems from the fact that the semantic concepts with the same modality can be extracted from both images and descriptions. To further improve the quality of captions generated by the model, we propose the Semantic Relationship Explorer, which explores the relationships between semantic concepts for better understanding of the image. Extensive experiments on MSCOCO dataset show that we can generate desirable captions without paired datasets. Furthermore, the proposed approach boosts five strong baselines under the paired setting, where the most significant improvement in CIDEr score reaches 8%, demonstrating that it is effective and generalizes well to a wide range of models.

cs.CV↗

arXiv2021

Personalized Graph Neural Networks with Attention Mechanism for Session-Aware Recommendation

The problem of session-aware recommendation aims to predict users' next click based on their current session and historical sessions. Existing session-aware recommendation methods have defects in capturing complex item transition relationships. Other than that, most of them fail to explicitly distinguish the effects of different historical sessions on the current session. To this end, we propose a novel method, named Personalized Graph Neural Networks with Attention Mechanism (A-PGNN) for brevity. A-PGNN mainly consists of two components: one is Personalized Graph Neural Network (PGNN), which is used to extract the personalized structural information in each user behavior graph, compared with the traditional Graph Neural Network (GNN) model, which considers the role of the user when the node embeddding is updated. The other is Dot-Product Attention mechanism, which draws on the Transformer net to explicitly model the effect of historical sessions on the current session. Extensive experiments conducted on two real-world data sets show that A-PGNN evidently outperforms the state-of-the-art personalized session-aware recommendation methods.

cs.IR↗

arXiv2019

Black and gray solitons in holographic superfluids at zero temperature

We construct gray soliton configurations, which move at constant speeds, in holographic superfluids for the first time. Since there should be no dissipation for a moving soliton to exist, we use the simplest holographic superfluid model at zero temperature, considering both the standard and alternative quantizations. For comparison purpose, we first investigate black solitons in the zero temperature holographic superfluids, which are static configurations. Then we focus on the numerical construction of gray solitons under both quantizations, which interpolate between the (static) black solitons and sound waves (moving at the speed of sound). Interestingly, under the standard quantization, a peculiar oscillation of the soliton configurations is observed, very much resembling the Friedel oscillation in fermionic superfluids at the BCS regime. Some implications and other aspects of the soliton configurations are also discussed.

hep-th↗

arXiv2014

Microwave control of trapped-ion motion assisted by a running optical lattice

We experimentally demonstrate microwave control of the motional state of a trapped ion placed in a state-dependent potential generated by a running optical lattice. Both the optical lattice depth and the running lattice frequency provide tunability of the spin-motion coupling strength. The spin-motional coupling is exploited to demonstrate sideband cooling of a Yb171 ion to the ground state of motion.

quant-ph↗

arXiv2010

Maximum relative excitation of a specific vibrational mode via optimum laser pulse duration

For molecules and materials responding to femtosecond-scale optical laser pulses, we predict maximum relative excitation of a Raman-active vibrational mode with period T when the pulse has an FWHM duration of 0.42 T. This result follows from a general analytical model, and is precisely confirmed by detailed density-functional-based dynamical simulations for C60 and a carbon nanotube, which include anharmonicity, nonlinearity, no assumptions about the polarizability tensor, and no averaging over rapid oscillations within the pulse. The mode specificity is, of course, best at low temperature and for pulses that are electronically off-resonance, and the energy deposited in any mode is proportional to the fourth power of the electric field.

cond-mat.mes-hall↗

arXiv2007

Itinerant and localized magnetism on the triangular lattice: sodium rich phases of Na$_x$CoO$_2$

We study the interplay between correlation, itinerant ferromagnetism and local moment formation on the electron doped triangular lattice of sodium cobaltates Na$_x$CoO$_2$. We find that strong correlation renormalizes the Stoner criterion and stabilizes the paramagnetic state for $x x_c$, ferromagnetic (FM) order emerges. The enhanced Na dopant potential fluctuations play a crucial role in the sodium rich phases and lead to an inhomogeneous FM state, exhibiting nonmagnetic Co$^{3+}$ patches, antiferromagnetic (AF) correlated regions, and FM clusters with AF domains. Hole doping the band insulator at x=1 leads to the formation of local moments near the Na vacancies and AF correlated magnetic clusters. We explain recent observations by neutron, $μ$SR, and NMR experiments on the evolution of the magnetic properties in the sodium rich phases.

cond-mat.str-el↗

arXiv2006

Vortex Properties of a Resonant Superfluid

The properties of a vortex in a rotating superfluid Fermi gas are studied in the unitary limit. A phenomenological approach based on Ginzburg-Landau theory is developed for this purpose. The density profiles, including those of the normal fluid and superfluid, are obtained at various temperatures and rotation frequencies. The superfluid and normal fluid densities can be identified from the angular momentum density. The total free energy and angular momentum of the vortex are also obtained.

cond-mat.other↗

arXiv2005

Electron correlation and Fermi surface topology of Na$_x$CoO$_2$

The electronic structure of Na$_x$CoO$_2$ revealed by recent photoemission experiments shows important deviations from band theory predictions. The six small Fermi surface pockets predicted by LDA calculations have not been observed as the associated $e_g^\prime$ band fails to cross the Fermi level for a wide range of sodium doping concentration $x$. In addition, significant bandwidth renormalizations of the $t_{2g}$ complex have been observed. We show that these discrepancies are due to strong electronic correlations by studying the multi-orbital Hubbard model in the Hartree-Fock and strong-coupling Gutzwiller approximation. The quasiparticle dispersion and the Fermi surface topology obtained in the presence of strong local Coulomb repulsion are in good agreement with experiments.

cond-mat.str-el↗