SearcharxivSearch

arXiv subjects

Aleksandr Nikitin

Publications and source records attributed to Aleksandr Nikitin.

2 recordsLinked to original sources

Oxygen-nonstoichiometry-driven phase transition in $\mathrm{Sr}_{1-x}\mathrm{Nd}_{x}\mathrm{CoO}_{3-\delta}$ ($x = 0.1, 0.2, 0.3$) perovskites

We report a systematic study of the interplay between oxygen nonstoichiometry, crystal structure, and magnetic/electrotransport properties in $\mathrm{Sr}_{1-x}\mathrm{Nd}_{x}\mathrm{CoO}_{3-\delta}$ ($x = 0.1, 0.2, 0.3$). High-resolution neutron powder diffraction combined with synchrotron x-ray powder diffraction reveals that increasing the oxygen content induces a structural transition from a layered $I4/mmm$ ($2a_p \times 2a_p \times 4a_p$) to an oxygen-deficient orthorhombic $Pmmm$ ($a_p \times a_p \times 2a_p$) phases with preferential oxygen-vacancy occupation. This transition is accompanied by a crossover from G-type antiferromagnetic with a weak ferromagnetic component to a ferromagnetic state, and a drastic decay in resistivity. The evolution of the magnetic and transport properties is discussed in terms of changes in the Co spin state, enhanced Co $3d$ - O $2p$ orbital overlap upon oxygen uptake, and a magnetically inhomogeneous ferromagnetic state associated with residual oxygen vacancies and mixed $\mathrm{Co}^{3+}/\mathrm{Co}^{4+}$ valence. Our findings experimentally confirm that the stabilization of the layered "314" structure is driven by the presence and ordering of oxygen vacancies rather than A-site cation ordering, whereas the oxygen-deficient oxidized compounds represent an intermediate orthorhombic state preceding fully stoichiometric phases.

cond-mat.mtrl-sci

Cryfish: On deep audio analysis with Large Language Models

The recent revolutionary progress in text-based large language models (LLMs) has contributed to the growth of interest in extending capabilities of such models to multimodal perception and understanding tasks. Hearing is an essential capability that is highly desired to be integrated into LLMs. However, effective integrating listening capabilities into LLMs is a significant challenge lying in generalizing complex auditory tasks across speech and sounds. To address these issues, we introduce Cryfish, our version of auditory-capable LLM. The model integrates WavLM audio-encoder features into Qwen2 model using a transformer-based connector. Cryfish is adapted to various auditory tasks through a specialized training strategy. We evaluate the model on the new Dynamic SUPERB Phase-2 comprehensive multitask benchmark specifically designed for auditory-capable models. The paper presents an in-depth analysis and detailed comparison of Cryfish with the publicly available models.

eess.AS