Explore the vast range of titles available.

High-quality test cases are vital for ensuring software reliability and security. However, existing symbolic execution tools generally rely on single-path search strategies, have limited feature extraction capability, and exhibit unstable model predictions. These limitations make them prone to local optima in complex or cross-scenario tasks and hinder their ability to balance testing quality with execution efficiency. To address these challenges, this paper proposes a Deep Active Ensemble Learning Framework for symbolic execution path exploration. During training, the framework integrates active learning with ensemble learning to reduce annotation costs and improve model robustness, while constructing a heterogeneous model pool to leverage complementary model strengths. In the testing stage, a dynamic ensemble mechanism based on sample similarity adaptively selects the optimal predictive model to guide symbolic path exploration. In addition, a gated graph neural network is employed to extract structural and semantic features from the control flow graph, improving program behavior understanding. To balance efficiency and coverage, a dynamic sliding window mechanism based on branch density enables real-time window adjustment under path complexity awareness. Experimental results on multiple real-world benchmark programs show that the proposed framework detects up to 16 vulnerabilities and achieves a cumulative 27.5% increase in discovered execution paths in hybrid fuzzing. Furthermore, the dynamic sliding window mechanism raises the F1 score to 93%.

Fluid therapy is central to sepsis management, yet recommendations on fluid type, volume, optimization, and de-escalation remain uncertain. The 2025 ESICM guidelines highlight major evidence gaps in sepsis fluid therapy. Although large language models (LLMs) show promise for guideline interpretation and clinical decision support, their performance in this high-risk domain is unclear. We conducted a prospective, cross-sectional observational study using nine guideline-derived sepsis-related clinical questions addressing fluid selection, resuscitation volume, and fluid removal during de-escalation. Questions were queried in both English and Chinese across three consecutive days, generating three independent responses per model from ChatGPT-5, ChatGPT-4o, and DeepSeek-V3.1. Three blinded intensivists evaluated responses for accuracy, completeness, and consistency using 5-point Likert scales. Readability was assessed using Flesch Reading Ease (FRE) and Flesch-Kincaid Grade Level (FKGL) for English responses and a validated Chinese readability framework. Inter-rater agreement was quantified using Kendall's W coefficient. In English responses, ChatGPT-5 achieved the highest accuracy, although inter-model differences were not statistically significant. In Chinese responses, ChatGPT-5 demonstrated significantly higher accuracy than ChatGPT-4o ( < 0.05). DeepSeek-V3.1 produced significantly more complete English responses than ChatGPT-4o ( < 0.05). Consistency was high across all models and languages. FKGL scores differed significantly among models ( < 0.01), with ChatGPT-5 generating more linguistically complex English text. No significant differences were observed between English and Chinese responses across evaluation dimensions. Advanced LLMs show potential for supporting sepsis fluid therapy guideline interpretation, but consistent overconfident responses in guideline-defined uncertainty domains highlight important safety limitations. Clinical oversight remains essential when deploying LLMs for high-risk decision support.

Quantum cascade lasers subject to tilted optical feedback produce periodic oscillations, quasi-periodic oscillations, and low-frequency oscillations. This work presents the modulation characteristics of period-one (P1) oscillations in a quantum cascade laser with tilted optical feedback. The electrical signal at the oscillation frequency is more than 50 dB higher than the noise level, and the electrical linewidth is less than 2.0 kHz. This electrical linewidth is about four orders of magnitude narrower than the optical linewidth (around 16 MHz) of the free-running laser, which suggests that the optical sidebands induced by the P1 oscillations are highly coherent with the main optical mode. In addition, the modulation depth of the optical signal is found to be in the range of 1% to 3.5%. In addition, it is verified in the simulations that the P1 oscillations induce not only amplitude modulation but also frequency modulation due to the phase-amplitude coupling effect.

The structural components of high speed trains are usually made of aluminum alloys, for example, 6082. The dynamic mechanical behavior of the material is one of key factors considered in structural design and safety assessment. In this paper, dynamic mechanical experiments were conducted with strain rate ranging from 0.001 s−1 to 100 s−1 using Instron tensile testing machine. The true stress-strain curves were fitted based on experimental data. Johnson-Cook model of 6082-T6 aluminum alloy was built to investigate the effect of strain and strain rate on flow stress. It has shown that the flow stress was sensitive to the strain rate. Yield strength and tensile strength increased with a high strain rate, which showed strain rate effect to some extent. Fracture analysis was carried out by using Backscattered Electron imaging (BSE). As strain rate increased, more precipitates were generated in fracture.

In recent years, the number of smart contracts deployed on blockchain has exploded. However, the issue of vulnerability has caused incalculable losses. Due to the irreversible and immutability of smart contracts, vulnerability detection has become particularly important. With the popular use of neural network model, there has been a growing utilization of deep learning-based methods and tools for the identification of vulnerabilities within smart contracts. This paper commences by providing a succinct overview of prevalent categories of vulnerabilities found in smart contracts. Subsequently, it categorizes and presents an overview of contemporary deep learning-based tools developed for smart contract detection. These tools are categorized based on their open-source status, the data format and the type of feature extraction they employ. Then we conduct a comprehensive comparative analysis of these tools, selecting representative tools for experimental validation and comparing them with traditional tools in terms of detection coverage and accuracy. Finally, Based on the insights gained from the experimental results and the current state of research in the field of smart contract vulnerability detection tools, we suppose to provide a reference standard for developers of contract vulnerability detection tools. Meanwhile, forward-looking research directions are also proposed for deep learning-based smart contract vulnerability detection.

Despite the early development of children’s sensitivity to the informativeness of testimony, there is limited understanding of their interpretation of others’ history of informativeness. This study investigates how preschoolers make trait inferences and pragmatic inferences about informants who differed in informativeness, and how these abilities affect their selective learning. Four- and 5-year-olds ( N = 64) observed two informants with differential access to a series of conjunctive causal events (full vs. partial). They were then asked to make pragmatic and trait inferences about the informants before choosing one informant to learn from. Five-year-olds, but not 4-year-olds, preferred to learn from the informative speaker. This pattern of selective learning held only for children who evaluated the informative speaker as smarter and for those children who could infer the informants’ epistemic states from the strength of statements. These findings highlight the crucial role of trait reasoning and pragmatic ability in guiding children’s selective learning.

Ensuring the absence of vulnerabilities or flaws in smart contracts before their deployment is crucial for the smooth progress of subsequent work. Existing detection methods heavily rely on expert rules, resulting in low robustness and accuracy. Therefore, we propose EDSCVD, an enhanced deep learning vulnerability detection model based on dual-channel networks. Firstly, the contract fragments are preprocessed by BERT into the required word embeddings. Next, we utilized adversarial training FGM to the word embeddings to generate perturbations, thereby producing symmetric adversarial samples and enhancing the robustness of the model. Then, the dual-channel model combining BiLSTM and CNN is utilized for feature training to obtain more comprehensive and symmetric information on temporal and local contract features.Finally, the combined output features are passed through a classifier to classify and detect contract vulnerabilities. Experimental results show that our EDSCVD exhibits excellent detection performance in the detection of classical reentrancy vulnerabilities, timestamp dependencies, and integer overflow vulnerabilities.

Novel fluorescent silver nanoclusters (AgNCs) were synthesized using amido black 10B (AB) as a stabilizing agent and then employed for the detection of copper ions (Cu 2+ ). The AB-stabilized AgNCs (AB–AgNCs) were well dispersed in aqueous solution with an average diameter of around 1.3 nm and exhibited illustrious blue fluorescence emission. Moreover, the fluorescence of AB–AgNCs could be quenched efficiently by Cu 2+ , which might be a result of the coordination between Cu 2+ and the free recognition group of AB on surfaces of AB–AgNCs, inducing the aggregation of AB–AgNCs. Based on the linear decrease of fluorescence intensity, the Cu 2+ concentration was determined in the range of 0.01–1.1 μmol L −1 and the limit of detection (LOD) was 4.0 nΜ. In addition, the detection of Cu 2+ could be performed with AB–AgNCs in the presence of other ions, including 13 kinds of conventional metal ions and 11 kinds of anions. Based on the above experiment, the developed AB–AgNC probe was successfully further applied to detect Cu 2+ in three electroplating effluents, which showed high accuracy. Graphical abstract The process of synthesised silver nanoclusters and the application for Cu 2+ detection.

Humanoid control systems have made significant progress in recent years, yet modeling fluent interaction-rich behavior between a robot, its surrounding environment, and task-relevant objects remains a fundamental challenge. This difficulty arises from the need to jointly capture spatial context, temporal dynamics, robot actions, and task intent at scale, which is a poor match to conventional supervision. We propose ExoActor, a novel framework that leverages the generalization capabilities of large-scale video generation models to address this problem. The key insight in ExoActor is to use third-person video generation as a unified interface for modeling interaction dynamics. Given a task instruction and scene context, ExoActor synthesizes plausible execution processes that implicitly encode coordinated interactions between robot, environment, and objects. Such video output is then transformed into executable humanoid behaviors through a pipeline that estimates human motion and executes it via a general motion controller, yielding a task-conditioned behavior sequence. To validate the proposed framework, we implement it as an end-to-end system and demonstrate its generalization to new scenarios without additional real-world data collection. Furthermore, we conclude by discussing limitations of the current implementation and outlining promising directions for future research, illustrating how ExoActor provides a scalable approach to modeling interaction-rich humanoid behaviors, potentially opening a new avenue for generative models to advance general-purpose humanoid intelligence.

Despite the rapid advancements in Multimodal Large Language Models (MLLMs), a critical question regarding their visual grounding mechanism remains unanswered: do these models genuinely ``read'' text embedded in images, or do they merely rely on parametric shortcuts in the text prompt? In this work, we diagnose this issue by introducing the Visualized-Question (VQ) setting, where text queries are rendered directly onto images to structurally mandate visual engagement. Our diagnostic experiments on Qwen2.5-VL reveal a startling capability-utilization gap: despite possessing strong OCR capabilities, models suffer a performance degradation of up to 12.7% in the VQ setting, exposing a deep-seated ``modality laziness.'' To bridge this gap, we propose SimpleOCR, a plug-and-play training strategy that imposes a structural constraint on the learning process. By transforming training samples into the VQ format with randomized styles, SimpleOCR effectively invalidates text-based shortcuts, compelling the model to activate and optimize its visual text extraction pathways. Empirically, SimpleOCR yields robust gains without architectural modifications. On four representative OOD benchmarks, it surpasses the base model by 5.4% and GRPO based on original images by 2.7%, while exhibiting extreme data efficiency, achieving superior performance with 30x fewer samples (8.5K) than recent RL-based methods. Furthermore, its plug-and-play nature allows seamless integration with advanced RL strategies like NoisyRollout to yield complementary improvements. Code is available at https://github.com/aiming-lab/SimpleOCR.