Existing benchmarks in e-commerce primarily focus on basic user intents, such as finding or purchasing products. However, real-world users often pursue more complex goals, such as applying vouchers, managing budgets, and finding multi-products seller. To bridge this gap, we propose ShoppingBench, a novel end-to-end shopping benchmark designed to encompass increasingly challenging levels of grounded intent. Specifically, we propose a scalable framework to simulate user instructions based on various intents derived from sampled real-world products. To facilitate consistent and reliable evaluations, we provide a large-scale shopping sandbox that serves as an interactive simulated environment, incorporating over 2.5 million real-world products. Experimental results demonstrate that even state-of-the-art language agents (such as GPT-4.1) achieve absolute success rates under 50% on our benchmark tasks, highlighting the significant challenges posed by our ShoppingBench. In addition, we propose a trajectory distillation strategy and leverage supervised fine-tuning, along with reinforcement learning on synthetic trajectories, to distill the capabilities of a large language agent into a smaller one. As a result, our trained agent achieves competitive performance compared to GPT-4.1.

Stochastic computing (SC) generally suffers from long latency. One solution is to apply proper random number sources (RNSs) to generate the bit streams. However, existing RNS designs either have low accuracy or high hardware cost. To address this drawback, motivated by the fact that a uniform spatial distribution generally leads to high accuracy for an SC circuit, we propose a basic architecture to produce a uniform spatial distribution and a further detailed implementation of it. For the implementation, we further propose a method to optimize its hardware cost and an algorithm following a guiding principle to improve its accuracy. The method for hardware cost optimization allows hardware cost reduction while keeping the accuracy. Our experimental results show that the proposed implementation achieves both high accuracy and low hardware cost. For example, compared to a state-of-the-art stochastic number generator design, our design can reduce hardware cost by over 80%, while achieving higher accuracy

Mining area scene classification is crucial for deposit evaluation and environmental monitoring. However, existing methods struggle with homogeneous and heterogeneous spectral spatial and topographic features of mining areas, large intra-class variations, and small target sizes. To overcome these limitations, this study integrates RGB and SAR data to construct a multi-modal dataset and proposes an RGB-SAR mining scene classification model with dual feature enhancement and adaptive cross-modal attention interaction. The model includes: (1) Dual feature enhancement module that suppresses irrelevant features and enhances discriminative multi-scale representations of mining targets; (2) BifocalNet based feature extraction module using a CNN-Transformer hybrid architecture to capture local textures and model global context; (3) Attention based adaptive cross-modal interaction module that achieves deep spectral geometric feature complementarity through the fusion of RGB and SAR modalities. Experiments show the model achieves an OA of 84.58%, outperforming other models and ranking first or second in most evaluation metrics. The proposed dataset and model thus advance mining scene classification.

Accurate identification of somatic small variants in tumors plays a crucial role in cancer diagnosis. Various somatic mutation callers have been developed; however, existing methods face limitations in modeling mapping information of flanking genomic sites (genomic sites adjacent to a somatic site) that influence the state of the somatic site. Additionally, they are unable to analyze inter-site interactions within the context sequence centered around a somatic site or appropriately weigh the effects of flanking genomic sites on the somatic site. To address these limitations, the Transformer model is utilized to develop TransSSVs for detecting somatic small variants. The core functionality of TransSSVs relies on the multi-head attention mechanism, which generates a reliable representation of interactions between a candidate somatic site and its flanking genomic sites within the context sequence. TransSSVs effectively extract mapping features of various genomic sites in the context sequence to enhance prediction accuracy. Benchmarking experiments demonstrate that TransSSVs exhibit robust performance when compared with state-of-the-art methods on well-characterized real and simulated tumor datasets. Furthermore, the contributions of flanking genomic sites to the detection of somatic sites are assessed, and attention weight patterns for positive and negative somatic sites are analyzed.

Influenza poses a significant global public health threat, with vaccination being the most effective and economical preventive measure. However, these punctuated antigenic changes, particularly in HA, result in escape from the immunity that was induced by prior infection or vaccination. Accurately predicting antigenic variation and understanding the antigenic dynamics of influenza viruses are crucial for selecting appropriate vaccine strains, but no established methods exist for influenza B viruses. Therefore, we present PREDAC-FluB, a hybrid deep learning framework that integrates spatial feature extraction via CNN to model interactions in HA1 sequences, multimodal sequence representation combining ESM-2 embeddings with six physicochemical descriptors and continuous encoding (ESM2-7-features), and UMAP-guided clustering for antigenic cluster identification. Using data from 9036 B/Victoria-lineage and 4520 B/Yamagata-lineage influenza virus pair. PREDAC-FluB demonstrates superior performance over traditional machine learning methods in predicting antigenic variation in influenza viruses, successfully identifying major antigenic clusters. Specifically, PREDAC-FluB classified the B/Victoria lineage into nine antigenic clusters and the B/Yamagata lineage into three antigenic clusters. In five-fold cross-validation for B/Victoria viruses, PREDAC-FluB with ESM2-7-features encoding achieved AUROC values of 0.9961 on the validation set and 0.9856 on the independent test set. In retrospective testing for B/Victoria viruses, PREDAC-FluB achieved AUROC values ranging from 0.83 to 0.97, demonstrating high prediction accuracy and effectively capturing antigenic variation information. In conclusion, PREDAC-FluB is a robust tool for antigenic computation, capable of accurately predicting antigenic variation in influenza B viruses. Its high prediction accuracy makes it a promising auxiliary method for recommending future influenza vaccine strains.

Stochastic computing (SC) is a promising computing paradigm for circuit design in the post-Moore era. It encodes data through stochastic bit streams (SBSs) and employs a randomizer to generate them, where the randomizer converts binary-encoded variables into stochastic formats and can optionally provide some SBSs of constant values. Owing to this, the randomizer generally plays a critical role in determining the accuracy of SC circuits and occupies a significant portion of their hardware cost. Therefore, it is crucial to apply proper randomizers to enhance the overall performance and efficiency of SC circuits. However, recent SC circuit designs often suffer from complex randomizers to ensure high accuracy. To address this issue, several efficient designs and optimization methods of randomizers have been proposed. In this paper, we review the common designs, the optimization methods, and the efficient application of randomizers, while discussing current challenges and future directions. By providing a brief overview, this survey underscores the critical role of randomizer design and optimization for efficient SC.

Vaccination stands as the most effective and economical strategy for prevention and control of influenza. The primary target of neutralizing antibodies is the surface antigen hemagglutinin (HA). However, ongoing mutations in the HA sequence result in antigenic drift. The success of a vaccine is contingent on its antigenic congruence with circulating strains. Thus, predicting antigenic variants and deducing antigenic clusters of influenza viruses are pivotal for recommendation of vaccine strains. The antigenicity of influenza A viruses is determined by the interplay of amino acids in the HA1 sequence. In this study, we exploit the ability of convolutional neural networks (CNNs) to extract spatial feature representations in the convolutional layers, which can discern interactions between amino acid sites. We introduce PREDAC-CNN, a model designed to track antigenic evolution of seasonal influenza A viruses. Accessible at http://predac-cnn.cloudna.cn, PREDAC-CNN formulates a spatially oriented representation of the HA1 sequence, optimized for the convolutional framework. It effectively probes interactions among amino acid sites in the HA1 sequence. Also, PREDAC-CNN focuses exclusively on physicochemical attributes crucial for the antigenicity of influenza viruses, thereby eliminating unnecessary amino acid embeddings. Together, PREDAC-CNN is adept at capturing interactions of amino acid sites within the HA1 sequence and examining the collective impact of point mutations on antigenic variation. Through 5-fold cross-validation and retrospective testing, PREDAC-CNN has shown superior performance in predicting antigenic variants compared to its counterparts. Additionally, PREDAC-CNN has been instrumental in identifying predominant antigenic clusters for A/H3N2 (1968-2023) and A/H1N1 (1977-2023) viruses, significantly aiding in vaccine strain recommendation.

Parkinson's Disease (PD) is a commonly diagnosed movement disorder with more than 10 million patients worldwide. Its clinical evaluation relies on a rating system called MDS-UPDRS, which includes subjective and error-prone motor examinations. This paper proposes an objective and interpretable visual system (PD-Net ) to quantitatively evaluate motor function of PD patients using video footage. The PD-Net consists of three modules: 1) a pose detector to infer 21 hand keypoints directly from RGB videos, 2) a movement analysis module to study temporal patterns of hand keypoints and discover motor symptoms, and 3) a scoring module to predict MDS-UPDRS ratings with retrieved symptoms. Trained with an in-house clinical dataset, PD-Net can effectively handle the unique challenges of PD examination videos, such as clinically-defined gestures, distinct self-occlusion/foreshortening effect and contextual background. And it detects hand keypoints of PD patients with an average accuracy of 84.1%, a 32.9% improvement over OpenPose. When compared to the ratings of experienced clinicians, PD-Net achieves an overall MDS-UPDRS rating score accuracy of 87.6% and Cohen's kappa of 0.82 on a testing dataset of 509 examination videos at a level exceeding human raters. This study demonstrates a clinically applicable automated video analysis system for PD clinical evaluation, which can facilitate early detection, routine monitoring, and treatment assessment.