Features Of Code Research Articles

Sparse graph cascade multi-kernel fusion contrastive learning for microbe–disease association prediction

Predicting microbe–disease associations (MDA) is crucial for proactively demystifying diseases causes and preventing them. Traditional prediction methods endure labor-intensive, time-consuming, and expensive. Therefore, this paper proposes CasMF-GCL, a novel Graph Contrastive Learning model based on sparse relation augmentation and a Cascaded Multi-kernel Fusion Network for MDA prediction. CasMF-GCL maximizes the use of sparse correlation information and eliminates data noise through an association augmentation technique based on low-rank sparse matrix completion. We employ diverse strategies to enhance the features of heterogeneous networks by exploring the similarity between microbes and diseases. Our model employs a new multi-layer graph convolutional network variant with a cascaded multi-kernel fusion mechanism, enabling weighted coding of local and global features for diseases and microbes. To further improve performance, we incorporate a self-supervised contrastive learning schema using multi-grain disease and microbe features and large-scale relation augmentations. Extensive experiments on two renowned datasets demonstrate that CasMF-GCL outperforms current state-of-the-art methods in seven indexes. The AUC values from 5-fold cross-validation are 0.997 and 0.989 for the two datasets, respectively. Ablation studies confirm the effectiveness of graph data augmentation, the power of the contrastive strategy, and the indispensability of the cascaded multi-kernel fusion network. Furthermore, case studies validate the prediction performance of CasMF-GCL.

EMPT: a sparsity Transformer for EEG-based motor imagery recognition.

Transformer network is widely emphasized and studied relying on its excellent performance. The self-attention mechanism finds a good solution for feature coding among multiple channels of electroencephalography (EEG) signals. However, using the self-attention mechanism to construct models on EEG data suffers from the problem of the large amount of data required and the complexity of the algorithm. We propose a Transformer neural network combined with the addition of Mixture of Experts (MoE) layer and ProbSparse Self-attention mechanism for decoding the time-frequency-spatial domain features from motor imagery (MI) EEG of spinal cord injury patients. The model is named as EEG MoE-Prob-Transformer (EMPT). The common spatial pattern and the modified s-transform method are employed for achieving the time-frequency-spatial features, which are used as feature embeddings to input the improved transformer neural network for feature reconstruction, and then rely on the expert model in the MoE layer for sparsity mapping, and finally output the results through the fully connected layer. EMPT achieves an accuracy of 95.24% on the MI EEG dataset for patients with spinal cord injury. EMPT has also achieved excellent results in comparative experiments with other state-of-the-art methods. The MoE layer and ProbSparse Self-attention inside the EMPT are subjected to visualisation experiments. The experiments prove that sparsity can be introduced to the Transformer neural network by introducing MoE and kullback-leibler divergence attention pooling mechanism, thereby enhancing its applicability on EEG datasets. A novel deep learning approach is presented for decoding EEG data based on MI.

Feasibility and Acceptability of a Multifaceted Observational Protocol to Investigate Lucidity in Advanced Dementia.

Episodes of lucidity (ELs), characterized by spontaneous, transient recovery of abilities, are reported across neurological conditions, including advanced dementia. Despite the significance of these events, existing research is limited to retrospective reports. Approaches to prospectively capturing and characterizing ELs in dementia are lacking. This pilot study determined the feasibility and acceptability of a multifaceted observational protocol to capture, characterize, and validate ELs in individuals with advanced dementia in hospice. Feasibility was measured through enrollment/retention rates, workload/usability assessment of video observation, and data processing procedures to facilitate the measurement of verbal and nonverbal features for EL characterization. The acceptability of observation and informant validation procedures was qualitatively examined with clinician and family caregiver participants. Study procedures were endorsed as highly acceptable among clinician (N = 49) and caregiver (N = 16) participants, demonstrating higher than anticipated thresholds for observation duration. Enrollment and retention rates for patient participants (N = 6) were 100% and 84%, respectively. Workload and usability measures demonstrated low disruptiveness and high ease of use after training. Longitudinal video observation among individuals with advanced dementia in hospice care for the detection of lucidity was feasible and highly acceptable. Multidimensional, temporal coding of features of ELs is feasible but time-intensive.

Visual-guided scene-aware audio generation method based on hierarchical feature codec and rendering decision

Visually guided spatial sound generation (VGSSG) is a well-suited multimodal learning method for dealing with recorded videos. However, existing methods are difficult to be directly applied to spatial sound generation for movie clips. This is mainly due to (1) the existence of Cinematic Audiovisual Language (CAL) in movies, which makes it difficult to construct spatial sound mapping models directly through data-driven based methods. (2) The problem of the inadequate model performance, which is caused by the excessive heterogeneous gap between audiovisual modal information. To solve the aforementioned problems, we propose a VGSSG method based on CAL decision-making and hierarchical feature coding and decoding, which effectively accomplishes spatial sound generation based on the CAL of movies. Specifically, to solve the problem of CAL modeling, a multimodal information-guided movie audio rendering decision maker is established, which can decide the rendering strategy based on the CAL of the current clip. To narrow the heterogeneous gap that hinders the fusion between audiovisual modal data, we propose a codec structure based on hierarchical fusion of audiovisual features and full-scale skip-connections, which improves the efficiency of the comprehensive utilization of audiovisual modal data, and demonstrates the effectiveness of adopting shallow features in VGSSG task. We integrate both 2-channel and 6-channel spatial audio generation into a unified framework. In addition, we establish a movie audiovisual bimodal dataset with hand-crafted CAL annotations. Experimentally, we demonstrate that compared with the existing methods, our method has higher performance in terms of reducing generation distortion.

METHOD FOR DETERMINING THE BIT GRID OVERFLOW OF A COMPUTER SYSTEM OPERATING IN THE SYSTEM OF RESIDUAL CLASSES

Context. Consideration of a set of examples of practical application of the procedure for identifying overflow of the bit grid of a computer system operating in a non-positional number system in residual classes. The object of the study is the process of processing data represented in the residual class system.
 Objective. The goal of the work is to consider and analyze examples of the bit grid overflow definition of a computer system when implementing the operation of adding two numbers in a system of residual classes based on the application of a method for determining the bit grid overflow, based on the use of the concept of number rank.
 Method. The specificity of the functioning of a computer system in a system of residual classes requires the implementation of not only modular operations, but also requires the implementation of additional, so-called non-modular operations. Non-modular operations include the operation of determining the overflow of the bit grid of a computer system in the system of residual classes. In a non-positional number system in residual classes, implementing the process of detecting overflow of the bit grid of a computer system is a difficult task to implement. The method considered in the work for determining the overflow of the bit grid is based on the use of positional features of a non-positional code of numbers in the system of residual classes, namely the true and calculated ranks of a number. The process of determining the overflow of the result of the operation of adding two numbers in the system of residual classes has been studied, since this arithmetic operation is the main, basic operation performed by a computer system.
 Results. The developed methods are justified theoretically and studied when performing arithmetic modular operations of addition, subtraction and multiplication using tabular procedures.
 Conclusions. The main advantage of the presented method is that the process of determining the overflow of the bit grid can be carried out in the dynamics of the computing process of the computer system, i.e. without stopping the solution of the problem. This circumstance makes it possible to reduce the unproductive expenditure of the computer system in the system of residual classes. In addition, this method can be used to control the operation of adding two numbers in the residual class system. This increases the reliability of obtaining the true result of the operation of adding two numbers in the system of residual classes.

Effective combining source code and opcode for accurate vulnerability detection of smart contracts in edge AI systems

Automating transactions using smart contracts extends the functionality of blockchains and secures the decentralization of blockchains in edge AI systems. Whereas, since plenty of smart contracts are deployed to support various decentralized edge applications, the security vulnerabilities of smart contracts will lead to huge irreversible losses. To deal with this problem, many deep learning-based methods have been developed for vulnerability detection. However, most existing methods use only contract source codes for feature extraction, resulting in low accuracy. In contrast, we propose a method based on deep learning model to integrate both the features of contract source codes and opcodes for vulnerability detection. Particularly, the contextual features are extracted based on opcodes while the expert pattern features are extracted from the source codes. Using the real-world dataset of Ethereum smart contracts targeting reentrancy vulnerability, experiment results demonstrate that our method outperforms the state-of-the-art methods and achieves 96.89% accuracy and 95.41% F1-Score.

A vulnerability severity prediction method based on bimodal data and multi-task learning

Facing the increasing number of software vulnerabilities, the automatic analysis of vulnerabilities has become an important task in the field of software security. However, the existing severity prediction methods are mainly based on vulnerability descriptions and ignore the relevant features of vulnerability code, which only includes unimodal information and result in low prediction accuracy. This paper proposes a vulnerability severity prediction method based on bimodal data and multi-task learning. First the bimodal data, which consists of the description and source code of each vulnerability, is preprocessed. Next the GraphCodeBert is used for the word embedding module to extract different vulnerability features from the bimodal data. Then the Bi-GRU with attention mechanism is adopted for further feature extraction of vulnerability severity. Considering the strong correlation between the two tasks of vulnerability severity prediction and exploitability prediction, this paper proposes a multi-task learning approach, which allows the model to learn the connection and shared information between different tasks through a hard parameter sharing strategy, so as to achieve more accurate and reliable prediction of vulnerability severity. Experimental results show that the severity prediction method proposed in this paper outperforms state-of-the-art methods, and can achieve an average F1 score of 93.83 % on the public vulnerability dataset.

PatchDPCC: A Patchwise Deep Compression Framework for Dynamic Point Clouds

When compressing point clouds, point-based deep learning models operate points in a continuous space, which has a chance to minimize the geometric fidelity loss introduced by voxelization in preprocessing. But these methods could hardly scale to inputs with arbitrary points. Furthermore, the point cloud frames are individually compressed, failing the conventional wisdom of leveraging inter-frame similarity. In this work, we propose a patchwise compression framework called patchDPCC, which consists of a patch group generation module and a point-based compression model. Algorithms are developed to generate patches from different frames representing the same object, and more importantly, these patches are regulated to have the same number of points. We also incorporate a feature transfer module in the compression model, which refines the feature quality by exploiting the inter-frame similarity. Our model generates point-wise features for entropy coding, which guarantees the reconstruction speed. The evaluation on the MPEG 8i dataset shows that our method improves the compression ratio by 47.01% and 85.22% when compared to PCGCv2 and V-PCC with the same reconstruction quality, which is 9% and 16% better than that D-DPCC does. Our method also achieves the fastest decoding speed among the learning-based compression models.

Plant mitochondrial genome map (PMGmap): A software tool for the comprehensive visualization of coding, noncoding and genome features of plant mitochondrial genomes.

Tools for visualizing genomes are essential for investigating genomic features and their interactions. Currently, tools designed originally for animal mitogenomes and plant plastomes are used to visualize the mitogens of plants but cannot accurately display features specific to plant mitogenomes, such as nonlinear exon arrangement for genes, the prevalence of functional noncoding features and complex chromosomal architecture. To address these problems, a software package, plant mitochondrial genome map (PMGmap), was developed using the Python programming language. PMGmap can draw genes at exon levels; draw cis- and trans-splicing gene maps, noncoding features and repetitive sequences; and scale genic regions by using the scaling of the genic regions on the mitogenome (SAGM) algorithm. It can also draw multiple chromosomes simultaneously. Compared with other state-of-the-art tools, PMGmap showed better performance in visualizing 405 plant mitogenomes, showing potential as an invaluable tool for plant mitogenome research. The web and container versions and the source code of PMGmap can be accessed through the following link: http://www.1kmpg.cn/pmgmap.

VulMPFF: A Vulnerability Detection Method for Fusing Code Features in Multiple Perspectives

Source code vulnerabilities are one of the significant threats to software security. Existing deep learning-based detection methods have proven their effectiveness. However, most of them extract code information on a single intermediate representation of code (IRC), which often fails to extract multiple information hidden in the code fully, significantly limiting their performance. To address this problem, we propose VulMPFF, a vulnerability detection method that fuses code features under multiple perspectives. It extracts IRC from three perspectives: code sequence, lexical and syntactic relations, and graph structure to capture the vulnerability information in the code, which effectively realizes the complementary information of multiple IRCs and improves vulnerability detection performance. Specifically, VulMPFF extracts serialized abstract syntax tree as IRC from code sequence, lexical and syntactic relation perspective, and code property graph as IRC from graph structure perspective, and uses Bi-LSTM model with attention mechanism and graph neural network with attention mechanism to learn the code features from multiple perspectives and fuse them to detect the vulnerabilities in the code, respectively. We design a dual-attention mechanism to highlight critical code information for vulnerability triggering and better accomplish the vulnerability detection task. We evaluate our approach on three datasets. Experiments show that VulMPFF outperforms existing state-of-the-art vulnerability detection methods (i.e., Rats, FlawFinder, VulDeePecker, SySeVR, Devign, and Reveal) in Acc and F1 score, with improvements ranging from 14.71% to 145.78% and 152.08% to 344.77%, respectively. Meanwhile, experiments in the open-source project demonstrate that VulMPFF has the potential to detect vulnerabilities in real-world environments.

Development of a mathematical model and algorithm for the sequence of fault detection of the ship's hull

The labor intensity of the hull fault detection process, which generally consists of work performed directly on the vessel and the preparation of the final report, can be reduced through the use of modern methods of computer data analysis. The sequence of measurements of defects on the vessel (residual thicknesses, deformations, cement and other sealings, cracks, holes) can be justified by solving a modified traveling salesman problem, reflecting the optimization of the path between points in space, with the goal function in the form of the spent finite time, which is the sum of the time for performing the measurements themselves, moving between the points of such measurements, various technical works, for example, installing a ladder for access to the area located above human growth and its movement to the right sector of work. The article develops a mathematical model of human motion in the space around the vessel, an algorithm of movement. The adaptation of the genetic algorithm for the problem under consideration is given, including the features of solution coding, crossing and mutation of solutions, and choosing the best option. The results of the test debugging simulation for the bow of the river pusher hull showed the operability of the developed algorithm and the possibility of its application for research and practical tasks.

Improvements in Neoplasm Classification in the International Classification of Diseases, Eleventh Revision: Systematic Comparative Study With the Chinese Clinical Modification of the International Classification of Diseases, Tenth Revision.

The International Classification of Diseases, Eleventh Revision (ICD-11) improved neoplasm classification. We aimed to study the alterations in the ICD-11 compared to the Chinese Clinical Modification of the International Classification of Diseases, Tenth Revision (ICD-10-CCM) for neoplasm classification and to provide evidence supporting the transition to the ICD-11. We downloaded public data files from the World Health Organization and the National Health Commission of the People's Republic of China. The ICD-10-CCM neoplasm codes were manually recoded with the ICD-11 coding tool, and an ICD-10-CCM/ICD-11 mapping table was generated. The existing files and the ICD-10-CCM/ICD-11 mapping table were used to compare the coding, classification, and expression features of neoplasms between the ICD-10-CCM and ICD-11. The ICD-11 coding structure for neoplasms has dramatically changed. It provides advantages in coding granularity, coding capacity, and expression flexibility. In total, 27.4% (207/755) of ICD-10 codes and 38% (1359/3576) of ICD-10-CCM codes underwent grouping changes, which was a significantly different change (χ21=30.3; P<.001). Notably, 67.8% (2424/3576) of ICD-10-CCM codes could be fully represented by ICD-11 codes. Another 7% (252/3576) could be fully described by uniform resource identifiers. The ICD-11 had a significant difference in expression ability among the 4 ICD-10-CCM groups (χ23=93.7; P<.001), as well as a considerable difference between the changed and unchanged groups (χ21=74.7; P<.001). Expression ability negatively correlated with grouping changes (r=-.144; P<.001). In the ICD-10-CCM/ICD-11 mapping table, 60.5% (2164/3576) of codes were postcoordinated. The top 3 postcoordinated results were specific anatomy (1907/3576, 53.3%), histopathology (201/3576, 5.6%), and alternative severity 2 (70/3576, 2%). The expression ability of postcoordination was not fully reflected. The ICD-11 includes many improvements in neoplasm classification, especially the new coding system, improved expression ability, and good semantic interoperability. The transition to the ICD-11 will inevitably bring challenges for clinicians, coders, policy makers and IT technicians, and many preparations will be necessary.

Efficient Spiking Neural Networks with Biologically Similar Lithium-Ion Memristor Neurons.

Benefiting from the brain-inspired event-driven feature and asynchronous sparse coding approach, spiking neural networks (SNNs) are becoming a potentially energy-efficient replacement for conventional artificial neural networks. However, neuromorphic devices used to construct SNNs persistently result in considerable energy consumption owing to the absence of sufficient biological parallels. Drawing inspiration from the transport nature of Na+ and K+ in synapses, here, a Li-based memristor (LixAlOy) was proposed to emulate the biological synapse, leveraging the similarity of Li as a homologous main group element to Na and K. The Li-based memristor exhibits ∼8 ns ultrafast operating speed, 1.91 and 0.72 linearity conductance modulation, and reproducible switching behavior, enabled by lithium vacancies forming a conductive filament mechanism. Moreover, these memristors are capable of simulating fundamental behaviors of a biological synapse, including long-term potentiation and long-term depression behaviors. Most importantly, a threshold-tunable leaky integrate-and-fire (TT-LIF) neuron is built using LixAlOy memristors, successfully integrating synaptic signals from both temporal and spatial levels and achieving an optimal threshold of SNNs. A computationally efficient TT-LIF-based SNN algorithm is also implemented for image recognition schemes, featuring a high recognition rate of 90.1% and an ultralow firing rate of 0.335%, which is 4 times lower than those of other memristor-based SNNs. Our studies reveal the ion dynamics mechanism of the LixAlOy memristor and confirm its potential in rapid switching and the construction of SNNs.

Encoding of Visual Objects in the Human Medial Temporal Lobe.

The human medial temporal lobe (MTL) plays a crucial role in recognizing visual objects, a key cognitive function that relies on the formation of semantic representations. Nonetheless, it remains unknown how visual information of general objects is translated into semantic representations in the MTL. Furthermore, the debate about whether the human MTL is involved in perception has endured for a long time. To address these questions, we investigated three distinct models of neural object coding-semantic coding, axis-based feature coding, and region-based feature coding-in each subregion of the human MTL, using high-resolution fMRI in two male and six female participants. Our findings revealed the presence of semantic coding throughout the MTL, with a higher prevalence observed in the parahippocampal cortex (PHC) and perirhinal cortex (PRC), while axis coding and region coding were primarily observed in the earlier regions of the MTL. Moreover, we demonstrated that voxels exhibiting axis coding supported the transition to region coding and contained information relevant to semantic coding. Together, by providing a detailed characterization of neural object coding schemes and offering a comprehensive summary of visual coding information for each MTL subregion, our results not only emphasize a clear role of the MTL in perceptual processing but also shed light on the translation of perception-driven representations of visual features into memory-driven representations of semantics along the MTL processing pathway.

Representation of Natural Contours by a Neural Population in Monkey V4.

The cortical visual area, V4, has been considered to code contours that contribute to the intermediate-level representation of objects. The neural responses to the complex contour features intrinsic to natural contours are expected to clarify the essence of the representation. To approach the cortical coding of natural contours, we investigated the simultaneous coding of multiple contour features in monkey (Macaca fuscata) V4 neurons and their population-level representation. A substantial number of neurons showed significant tuning for two or more features such as curvature and closure, indicating that a substantial number of V4 neurons simultaneously code multiple contour features. A large portion of the neurons responded vigorously to acutely curved contours that surrounded the center of classical receptive field, suggesting that V4 neurons tend to code prominent features of object contours. The analysis of mutual information (MI) between the neural responses and each contour feature showed that most neurons exhibited similar magnitudes for each type of MI, indicating that many neurons showing the responses depended on multiple contour features. We next examined the population-level representation by using multidimensional scaling analysis. The neural preferences to the multiple contour features and that to natural stimuli compared with silhouette stimuli increased along with the primary and secondary axes, respectively, indicating the contribution of the multiple contour features and surface textures in the population responses. Our analyses suggested that V4 neurons simultaneously code multiple contour features in natural images and represent contour and surface properties in population.

Vulnerable JavaScript functions detection using stacking of convolutional neural networks.

System security for web-based applications is paramount, and for the avoidance of possible cyberattacks it is important to detect vulnerable JavaScript functions. Developers and security analysts have long relied upon static analysis to investigate vulnerabilities and faults within programs. Static analysis tools are used for analyzing a program's source code and identifying sections of code that need to be further examined by a human analyst. This article suggests a new approach for identifying vulnerable code in JavaScript programs by using ensemble of convolutional neural networks (CNNs) models. These models use vulnerable information and code features to detect related vulnerable code. For identifying different vulnerabilities in JavaScript functions, an approach has been tested which involves the stacking of CNNs with misbalancing, random under sampler, and random over sampler. Our approach uses these CNNs to detect vulnerable code and improve upon current techniques' limitations. Previous research has introduced several approaches to identify vulnerable code in JavaScript programs, but often have their own limitations such as low accuracy rates and high false-positive or false-negative results. Our approach addresses this by using the power of convolutional neural networks and is proven to be highly effective in the detection of vulnerable functions that could be used by cybercriminals. The stacked CNN approach has an approximately 98% accuracy, proving its robustness and usability in real-world scenarios. To evaluate its efficacy, the proposed method is trained using publicly available JavaScript blocks, and the results are assessed using various performance metrics. The research offers a valuable insight into better ways to protect web-based applications and systems from potential threats, leading to a safer online environment for all.

Analyzing and Discovering Spatial Algorithm Complexity Vulnerabilities in Recursion

The algorithmic complexity vulnerability (ACV) that may lead to denial of service attacks greatly disrupts the security and availability of applications, and due to the widespread use of third-party libraries, its impact may be amplified through the software supply chain. The existing work in the field is dedicated to abstract loop or iterative patterns and fuzzing the entire application to discover algorithm complexity vulnerabilities, but they still face efficiency and effectiveness issues. Our research focuses on: (1) proposing a representation and extraction method for code features related to algorithmic complexity vulnerabilities, helping analysts quickly understand program logic; (2) providing a new ACV detecting model, focusing on the spatial complexity anomalies caused by deep recursion structures, and proposing a new filtering method; and (3) aiming at the difficulty of efficiently generating complex-data-type-related payloads using existing symbol execution techniques, a call-chain-guided payload construction method is proposed. We tested third-party components in the open-source Java Maven Repository, identified many unexposed vulnerabilities, and eight of them received Common Vulnerabilities and Exposures (CVE) identifiers, and demonstrated that our method can discover more algorithmic complexity vulnerabilities compared to existing tools with better performance.

Predicting circRNA-RBP Binding Sites Using a Hybrid Deep Neural Network.

Circular RNAs (circRNAs) are non-coding RNAs generated by reverse splicing. They are involved in biological process and human diseases by interacting with specific RNA-binding proteins (RBPs). Due to traditional biological experiments being costly, computational methods have been proposed to predict the circRNA-RBP interaction. However, these methods have problems of single feature extraction. Therefore, we propose a novel model called circ-FHN, which utilizes only circRNA sequences to predict circRNA-RBP interactions. The circ-FHN approach involves feature coding and a hybrid deep learning model. Feature coding takes into account the physicochemical properties of circRNA sequences and employs four coding methods to extract sequence features. The hybrid deep structure comprises a convolutional neural network (CNN) and a bidirectional gated recurrent unit (BiGRU). The CNN learns high-level abstract features, while the BiGRU captures long-term dependencies in the sequence. To assess the effectiveness of circ-FHN, we compared it to other computational methods on 16 datasets and conducted ablation experiments. Additionally, we conducted motif analysis. The results demonstrate that circ-FHN exhibits exceptional performance and surpasses other methods. circ-FHN is freely available at https://github.com/zhaoqi106/circ-FHN .

Statement-Grained Hierarchy Enhanced Code Summarization

Code summarization plays a vital role in aiding developers with program comprehension by generating corresponding textual descriptions for code snippets. While recent approaches have concentrated on encoding the textual and structural characteristics of source code, they often neglect the global hierarchical features, causing limited code representation. Addressing this gap, our paper introduces the statement-grained hierarchy enhanced Transformer model (SHT), a novel framework that integrates global hierarchy, syntax, and token sequences to automatically generate summaries for code snippets. SHT is distinctively designed with two encoders to learn both hierarchical and sequential features of code. One relational attention encoder processes the statement-grained hierarchical graph, producing hierarchical embeddings. Subsequently, another sequence encoder integrates these hierarchical structures with token sequences. The resulting enriched representation is then fed into a vanilla Transformer decoder, which effectively generates concise and informative summarizations. Our extensive experiments demonstrate that SHT significantly outperforms state-of-the-art approaches on two widely used Java benchmarks. This underscores the effectiveness of incorporating global hierarchical information in enhancing the quality of code summarizations.

TOSCA – an open-source, finite-volume, large-eddy simulation (LES) environment for wind farm flows

Abstract. The growing number and growing size of wind energy projects coupled with the rapid growth in high-performance computing technology are driving researchers toward conducting large-scale simulations of the flow field surrounding entire wind farms. This requires highly parallel-efficient tools, given the large number of degrees of freedom involved in such simulations, and yields valuable insights into farm-scale physical phenomena, such as gravity wave interaction with the wind farm and farm–farm wake interactions. In the current study, we introduce the open-source, finite-volume, large-eddy simulation (LES) code TOSCA (Toolbox fOr Stratified Convective Atmospheres) and demonstrate its capabilities by simulating the flow around a finite-size wind farm immersed in a shallow, conventionally neutral boundary layer (CNBL), ultimately assessing gravity-wave-induced blockage effects. Turbulent inflow conditions are generated using a new hybrid off-line–concurrent-precursor method. Velocity is forced with a novel pressure controller that allows us to prescribe a desired average hub-height wind speed while avoiding inertial oscillations above the atmospheric boundary layer (ABL) caused by the Coriolis force, a known problem in wind farm LES studies. Moreover, to eliminate the dependency of the potential-temperature profile evolution on the code architecture observed in previous studies, we introduce a method that allows us to maintain the mean potential-temperature profile constant throughout the precursor simulation. Furthermore, we highlight that different codes do not predict the same velocity inside the boundary layer under geostrophic forcing owing to their intrinsically different numerical dissipation. The proposed methodology allows us to reduce such spread by ensuring that inflow conditions produced from different codes feature the same hub wind and thermal stratification, regardless of the adopted precursor run time. Finally, validation of actuator line and disk models, CNBL evolution, and velocity profiles inside a periodic wind farm is also presented to assess TOSCA’s ability to model large-scale wind farm flows accurately and with high parallel efficiency.