Restricted Boltzmann Machines Implemented by Spin-Orbit Torque Magnetic Tunnel Junctions.

Artificial intelligence has surged forward with the advent of generative models, which rely heavily on stochastic computing architectures enhanced by true random number generators with adjustable sampling probabilities. In this study, we develop spin-orbit torque magnetic tunnel junctions (SOT-MTJs), investigating their sigmoid-style switching probability as a function of the driving voltage. This feature proves to be ideally suited for stochastic computing algorithms such as the restricted Boltzmann machines (RBM) prevalent in pretraining processes. We exploit SOT-MTJs as both stochastic samplers and network nodes for RBMs, enabling the implementation of RBM-based neural networks to achieve recognition tasks for both handwritten and spoken digits. Moreover, we further harness the weights derived from the preceding image and speech training processes to facilitate cross-modal learning from speech to image generation. Our results clearly demonstrate that these SOT-MTJs are promising candidates for the development of hardware accelerators tailored for Boltzmann neural networks and other stochastic computing architectures.

Steering the Reaction Pathway of CO2 Electroreduction by Tuning the Coordination Number of Copper Catalysts.

Cu-based catalysts are optimal for the electroreduction of CO2 to generate hydrocarbon products. However, controlling product distribution remains a challenging topic. The theoretical investigations have revealed that the coordination number (CN) of Cu considerably influences the adsorption energy of *CO intermediates, thereby affecting the reaction pathway. Cu catalysts with different CNs were fabricated by reducing CuO precursors via cyclic voltammetry (Cyc-Cu), potentiostatic electrolysis (Pot-Cu), and pulsed electrolysis (Pul-Cu), respectively. High-CN Cu catalysts predominantly generate C2+ products, while low-CN Cu favors CH4 production. For instance, over the high-CN Pot-Cu, C2+ is the main product, with the Faradaic efficiency (FE) reaching 82.5% and a partial current density (j) of 514.3 mA cm-2. Conversely, the low-CN Pul(3)-Cu favors the production of CH4, achieving the highest FECH4 value of 56.7% with a jCH4 value of 234.4 mA cm-2. In situ X-ray absorption spectroscopy and Raman spectroscopy studies further confirm the different *CO adsorptions over Cu catalysts with different CN, thereby directing the reaction pathway of the CO2RR.

Highly Efficient Electrosynthesis of Glycine over an Atomically Dispersed Iron Catalyst.

Glycine is a nonessential amino acid that plays a vital role in various biological activities. However, the conventional synthesis of glycine requires sophisticated procedures or toxic feedstocks. Herein, we report an electrochemical pathway for glycine synthesis via the reductive coupling of oxalic acid and nitrate or nitrogen oxides over atomically dispersed Fe-N-C catalysts. A glycine selectivity of 70.7% is achieved over Fe-N-C-700 at -1.0 V versus RHE. Synergy between the FeN3C structure and pyrrolic nitrogen in Fe-N-C-700 facilitates the reduction of oxalic acid to glyoxylic acid, which is crucial for producing glyoxylic acid oxime and glycine, and the FeN3C structure could reduce the energy barrier of *HOOCCH2NH2 intermediate formation thus accelerating the glyoxylic acid oxime conversion to glycine. This new synthesis approach for value-added chemicals using simple carbon and nitrogen sources could provide sustainable routes for organonitrogen compound production.

Abnormal Degraded Tapetum 1 (ADT1) is required for tapetal cell death and pollen development in rice.

The timely degradation of tapetum, the innermost somatic anther cell layer in flowering plants, is critical for pollen development. Although several genes involved in tapetum development have been characterized, the molecular mechanisms underlying tapetum degeneration remain elusive. Here, we showed that mutation in Abnormal Degraded Tapetum 1 (ADT1) resulted in overaccumulation of Reactive Oxygen Species (ROS) and abnormal anther development, causing earlier tapetum Programmed Cell Death (PCD) and pollen abortion. ADT1 encodes a nuclear membrane localized protein, which is strongly expressed in the developing microspores and tapetal cells during early anther development. Moreover, ADT1 could interact with metallothionein MT2b, which was related to ROS scavenging and cell death regulation. These findings indicate that ADT1 is required for proper timing of tapetum PCD by regulating ROS homeostasis, expanding our understanding of the regulatory network of male reproductive development in rice.

Switching Reaction Pathways of CO2 Electroreduction by Modulating Cations in the Electrochemical Double Layer.

Tuning the selectivity of CO2 electroreduction reaction (CO2RR) solely by changing electrolyte is a very attractive topic. In this study, we conducted CO2RR in different aqueous electrolytes over bulk metal electrodes. It was discovered that controlled CO2RR could be achieved by modulating cations in the electrochemical double layer. Specifically, ionic liquid cations in the electrolyte significantly inhibits the hydrogen evolution reaction (HER), while yielding high Faraday efficiencies toward CO (FECO) or formate (FEformate) depending on the alkali metal cations. For example, the product could be switched from CO (FECO = 97.3%) to formate (FEformate = 93.5%) by changing the electrolyte from 0.1 M KBr-0.5 M 1-octyl-3-methylimidazolium bromide (OmimBr) to 0.1 M CsBr-0.5M OmimBr aqueous solutions over pristine Cu foil electrode. In situ spectroscopy and theoretical calculations reveal that the ordered structure generated by the assembly of Omim+ under an applied negative potential alters the hydrogen bonding structure of the interfacial water, thereby inhibiting the HER. The difference in selectivity in the presence of different cations is attributed to the hydrogen bonding effect caused by Omim+, which alters the solvated structure of the alkali metal cations and thus affects the stabilization of intermediates of different pathways.

LRFNet: A real-time medical image fusion method guided by detail information.

Multimodal medical image fusion (MMIF) technology plays a crucial role in medical diagnosis and treatment by integrating different images to obtain fusion images with comprehensive information. Deep learning-based fusion methods have demonstrated superior performance, but some of them still encounter challenges such as imbalanced retention of color and texture information and low fusion efficiency. To alleviate the above issues, this paper presents a real-time MMIF method, called a lightweight residual fusion network. First, a feature extraction framework with three branches is designed. Two independent branches are used to fully extract brightness and texture information. The fusion branch enables different modal information to be interactively fused at a shallow level, thereby better retaining brightness and texture information. Furthermore, a lightweight residual unit is designed to replace the conventional residual convolution in the model, thereby improving the fusion efficiency and reducing the overall model size by approximately 5 times. Finally, considering that the high-frequency image decomposed by the wavelet transform contains abundant edge and texture information, an adaptive strategy is proposed for assigning weights to the loss function based on the information content in the high-frequency image. This strategy effectively guides the model toward preserving intricate details. The experimental results on MRI and functional images demonstrate that the proposed method exhibits superior fusion performance and efficiency compared to alternative approaches. The code of LRFNet is available at https://github.com/HeDan-11/LRFNet.

Knowledge mapping of anaplastic thyroid cancer treatments: a bibliometric analysis (2000-2023).

Context: Anaplastic thyroid cancer (ATC) is a relatively rare and extensively malignant kind of thyroid carcinoma. The poor prognosis and high mortality rate of ATC can be attributed to its invasive features and undifferentiated phenotype. At present, there is a lack of efficacious therapeutic options. In light of the elevated fatality rate, it is vital to possess a comprehensive comprehension of the scientific terrain pertaining to ATC. To gather the perspectives of different researchers about the topic of ATC treatment, we did a bibliometric network analysis, which offers a comprehensive view of the scholarly literature. Methodology: A systematic search was conducted on the WoSCC database to identify publications pertaining to ATC treatment between the years 2000 and 2023. In this bibliometric investigation, the tools VOSviewers, CiteSpace, and the R package "bibliometrix" were employed to investigate the general attributes, developmental framework, and academic frontiers of the subject matter. Results: 1223 publications in total, written by 6937 scholars from 53 areas and 1402 institutions and published in 358 scholarly journals, were analyzed. There has been a gradual increase in the quantity of publications pertaining to ATC treatment. The United States and China emerged as the most prominent nations. The University of Texas MD Anderson Cancer Center and Memorial Sloan Kettering Cancer Counseling Center are prominent research institutions in highly productive countries. The journal Thyroid holds a prominent position within its discipline, being widely recognized as both the most popular and highly co-cited publication. According to the available data, Maria Cabanillas has authored the highest number of published articles, while RC Smallridge has received the highest number of co-citations. It turned out that the prevailing keywords encompassed expression, therapy, apoptosis, survival, activation, proliferation, metastasis, and other related terms. Immunotherapy, targeted therapy, and prognostic factors are the emerging research hotspots and trends. Conclusions: This paper presents a complete overview of research trends and advancements in the treatment of ATC using bibliometric analysis. The acquisition of information will offer vital insights for funding and potential creative strategies in researching the treatment of ATC, which indicates the research frontiers as well as prevalent directions in recent years.

A data-efficient and easy-to-use lip language interface based on wearable motion capture and speech movement reconstruction.

Lip language recognition urgently needs wearable and easy-to-use interfaces for interference-free and high-fidelity lip-reading acquisition and to develop accompanying data-efficient decoder-modeling methods. Existing solutions suffer from unreliable lip reading, are data hungry, and exhibit poor generalization. Here, we propose a wearable lip language decoding technology that enables interference-free and high-fidelity acquisition of lip movements and data-efficient recognition of fluent lip language based on wearable motion capture and continuous lip speech movement reconstruction. The method allows us to artificially generate any wanted continuous speech datasets from a very limited corpus of word samples from users. By using these artificial datasets to train the decoder, we achieve an average accuracy of 92.0% across individuals (n = 7) for actual continuous and fluent lip speech recognition for 93 English sentences, even observing no training burn on users because all training datasets are artificially generated. Our method greatly minimizes users' training/learning load and presents a data-efficient and easy-to-use paradigm for lip language recognition.

Single-Atom palladium engineered cobalt nanocomposite for selective aerobic oxidation of sulfides to sulfoxides.

Developing efficient catalysts for the selective oxidation of sulfides to sulfoxides using molecular oxygen as the oxidant is a challenging task. Here, we report a novel catalyst comprising a single atom palladium engineered cobalt nanocomposite (denoted as PdCo@NC-800-0.01) for this reaction. The incorporation of single atom palladium effectively transforms an originally inactive cobalt nanocomposite into a highly efficient and selective catalyst for the oxidation of sulfides. This catalyst PdCo@NC-800-0.01 exhibited outstanding performance in the selective oxidation of sulfides to sulfoxides using O2 as the oxidant in the presence of isobutyraldehyde (IBA) under mild conditions, demonstrating high activity and excellent selectivity for a broad spectrum of sulfides with good tolerance toward various functional groups, including those susceptible to oxidation. Furthermore, the catalyst could be easily recovered and reused up to 10 times without any significant loss in activity and selectivity. Comprehensive characterizations and theoretical calculations revealed that the engineering of cobalt nanocomposite with single atom Pd greatly enhanced the ability to adsorb and activate IBA, leading to the generation of the key acyl radical. This radical then reacted with singlet oxygen 1O2 derived from molecular oxygen, producing reactive oxygen species peroxy radical, which ultimately promoted the catalytic performance.

Probability-Distribution-Configurable True Random Number Generators Based on Spin-Orbit Torque Magnetic Tunnel Junctions.

The incorporation of randomness into stochastic computing can provide ample opportunities for applications such as simulated annealing, non-polynomial hard problem solving, and Bayesian neuron networks. In these cases, a considerable number of random numbers with an accurate and configurable probability distribution function (PDF) are indispensable. Preferably, these random numbers are provided at the hardware level to improve speed, efficiency, and parallelism. In this paper, how spin-orbit torque magnetic tunnel junctions (SOT-MTJs) with high barriers are suitable candidates for the desired true random number generators is demonstrated. Not only do these SOT-MTJs perform excellently in speed and endurance, but their randomness can also be conveniently and precisely controlled by a writing voltage, which makes them a well-performed Bernoulli bit. By utilizing these SOT-MTJ-based Bernoulli bits, any PDF, including Gaussian, uniform, exponential, Chi-square, and even arbitrarily defined distributions can be realized. These PDF-configurable true random number generators can then promise to advance the development of stochastic computing and broaden the applications of the SOT-MTJs.

All-in-one, all-optical logic gates using liquid metal plasmon nonlinearity.

Electronic processors are reaching the physical speed ceiling that heralds the era of optical processors. Multifunctional all-optical logic gates (AOLGs) of massively parallel processing are of great importance for large-scale integrated optical processors with speed far in excess of electronics, while are rather challenging due to limited operation bandwidth and multifunctional integration complexity. Here we for the first time experimentally demonstrate a reconfigurable all-in-one broadband AOLG that achieves nine fundamental Boolean logics in a single configuration, enabled by ultrabroadband (400-4000 nm) plasmon-enhanced thermo-optical nonlinearity (TONL) of liquid-metal Galinstan nanodroplet assemblies (GNAs). Due to the unique heterogeneity (broad-range geometry sizes, morphology, assembly profiles), the prepared GNAs exhibit broadband plasmonic opto-thermal effects (hybridization, local heating, energy transfer, etc.), resulting in a huge nonlinear refractive index under the order of 10-4-10-5 within visual-infrared range. Furthermore, a generalized control-signal light route is proposed for the dynamic TONL modulation of reversible spatial-phase shift, based on which nine logic functions are reconfigurable in one single AOLG configuration. Our work will provide a powerful strategy on large-bandwidth all-optical circuits for high-density data processing in the future.

Genome-wide identification and phylogenetic analysis of the ERF gene family in cucumbers

Members of the ERF transcription-factor family participate in a number of biological processes, viz., responses to hormones, adaptation to biotic and abiotic stress, metabolism regulation, beneficial symbiotic interactions, cell differentiation and developmental processes. So far, no tissue-expression profile of any cucumber ERF protein has been reported in detail. Recent completion of the cucumber full-genome sequence has come to facilitate, not only genome-wide analysis of ERF family members in cucumbers themselves, but also a comparative analysis with those in Arabidopsis and rice. In this study, 103 hypothetical ERF family genes in the cucumber genome were identified, phylogenetic analysis indicating their classification into 10 groups, designated I to X. Motif analysis further indicated that most of the conserved motifs outside the AP2/ERF domain, are selectively distributed among the specific clades in the phylogenetic tree. From chromosomal localization and genome distribution analysis, it appears that tandem-duplication may have contributed to CsERF gene expansion. Intron/exon structure analysis indicated that a few CsERFs still conserved the former intron-position patterns existent in the common ancestor of monocots and eudicots. Expression analysis revealed the widespread distribution of the cucumber ERF gene family within plant tissues, thereby implying the probability of their performing various roles therein. Furthermore, members of some groups presented mutually similar expression patterns that might be related to their phylogenetic groups.