Explore the vast range of titles available.

We experimentally studied the transverse mode instability (TMI) threshold of a linearly polarized single-frequency fiber laser amplifier constructed with tapered ytterbium-doped fiber (TYDF) under different bending diameters. The TMI threshold increased from 333 W to 451 W by reducing the bending diameter from 16 cm to 12 cm, which was accompanied by the deterioration of the beam quality from 1.47 to 1.67. The anomalous characteristics between the TMI threshold, bending diameter, and beam quality are mainly attributed to the decreased bending loss of higher-order mode (HOM) content as a result of the increased system heat loads caused by a tight bending-induced loss of amplification efficiency. It is believed that the presented results will provide useful guidelines for the design of high-power single-frequency fiber amplifiers.

Polylactic acid (PLA) materials face inherent limitations in many applications due to their low toughness. To address this challenge, this study employed a reactive melt-grafting method to prepare maleic anhydride (MA)-grafted poly(butylene adipate-co-terephthalate) (PBAT–MA), providing an effective approach to improve the interfacial compatibility between PLA and PBAT, thereby significantly enhancing the toughness and impact resistance of PLA and expanding its application scope. The grafting reaction process of PBAT–MA was investigated, as well as its toughening mechanism and effect on PLA. The results showed that at a maleic anhydride concentration of 2 wt%, the obtained PLA–PBAT–MA composite material exhibited the best performance, with a fracture elongation of 358.1%, 450.4% higher than that of the unmodified composite material. The impact strength was 333.9 kJ/m2, 917.3% higher than that of the unmodified composite material. This enhanced effect is attributed to the optimal MA concentration preserving the tough structure of PBAT while effectively bridging the interface between PLA and PBAT, promoting efficient stress transfer between the two phases, and ultimately achieving exceptional toughness.

Background Candida auris infections pose a threat to public health, necessitating increased awareness in China. This study aimed to analyze the strains of C. auris, assess the infection status, and investigate clinical characteristics and risk factors for mortality. Methods A retrospective analysis was conducted on 18 patients with Candida auris infection. We focused on evaluating basic characteristics, strain sources, and antibacterial susceptibility test results. Statistical methods were used to determine clinical features and identify risk factors for death. Results The strain type, composition ratio, and specimen source of C. auris were not associated with mortality. Neither the infection index nor the length of hospitalization showed an association with the prognosis. However, significant risk factors for mortality included cerebral infarction, cardiac disease, renal dysfunction, hypoproteinemia, and anemia (all p < 0.05). Conclusions Cerebral infarction, cardiac disease, renal dysfunction, hypoproteinemia, and anemia are significant risk factors for death in C. auris infections. These findings indicate the importance of recognizing and addressing these factors in the clinical management of C. auris infection. This research investigates the strains of C. auris, assesses the infection situation, and investigates clinical characteristics and risk factors for mortality. The significant risk factors for death in C. auris infections include (1) cerebral infarction (2) cardiac disease (3) renal dysfunction (4) hypoproteinemia (5) anemia.

Spatiotemporal reasoning plays a key role in Cyber-Physical Systems (CPS). Despite advances in Large Language Models (LLMs) and Large Reasoning Models (LRMs), their capacity to reason about complex spatiotemporal signals remains underexplored. This paper proposes a hierarchical SpatioTemporal reAsoning benchmaRK, STARK, to systematically evaluate LLMs across three levels of reasoning complexity: state estimation (e.g., predicting field variables, localizing and tracking events in space and time), spatiotemporal reasoning over states (e.g., inferring spatial-temporal relationships), and world-knowledge-aware reasoning that integrates contextual and domain knowledge (e.g., intent prediction, landmark-aware navigation). We curate 26 distinct spatiotemporal tasks with diverse sensor modalities, comprising 14,552 challenges where models answer directly or by Python Code Interpreter. Evaluating 3 LRMs and 8 LLMs, we find LLMs achieve limited success in tasks requiring geometric reasoning (e.g., multilateration or triangulation), particularly as complexity increases. Surprisingly, LRMs show robust performance across tasks with various levels of difficulty, often competing or surpassing traditional first-principle-based methods. Our results show that in reasoning tasks requiring world knowledge, the performance gap between LLMs and LRMs narrows, with some LLMs even surpassing LRMs. However, the LRM o3 model continues to achieve leading performance across all evaluated tasks, a result attributed primarily to the larger size of the reasoning models. STARK motivates future innovations in model architectures and reasoning paradigms for intelligent CPS by providing a structured framework to identify limitations in the spatiotemporal reasoning of LLMs and LRMs.

Practitioners deploying time series forecasting models face a dilemma: exhaustively validating dozens of models is computationally prohibitive, yet choosing the wrong model risks poor performance. We show that spectral predictability~\\(\\) -- a simple signal processing metric -- systematically stratifies model family performance, enabling fast model selection. We conduct controlled experiments in four different domains, then further expand our analysis to 51 models and 28 datasets from the GIFT-Eval benchmark. We find that large time series foundation models (TSFMs) systematically outperform lightweight task-trained baselines when \\(\\) is high, while their advantage vanishes as \\(\\) drops. Computing \\(\\) takes seconds per dataset, enabling practitioners to quickly assess whether their data suits TSFM approaches or whether simpler, cheaper models suffice. We demonstrate that \\(\\) stratifies model performance predictably, offering a practical first-pass filter that reduces validation costs while highlighting the need for models that excel on genuinely difficult (low-\\(\\)) problems rather than merely optimizing easy ones.