With the rise in sedentary behavior, health problems caused by poor sitting posture have drawn increasing attention. Most existing methods, whether using invasive sensors or computer vision, rely on two-stage pipelines, which result in high intrusiveness, intensive computation, and poor real-time performance on embedded edge devices. Inspired by YOLOv11-Pose, a lightweight single-stage network for sitting posture recognition on embedded edge devices termed LSP-YOLO was proposed. By integrating partial convolution(PConv) and Similarity-Aware Activation Module(SimAM), a lightweight module, Light-C3k2, was designed to reduce computational cost while maintaining feature extraction capability. In the recognition head, keypoints were directly mapped to posture classes through pointwise convolution, and intermediate supervision was employed to enable efficient fusion of pose estimation and classification. Furthermore, a dataset containing 5,000 images across six posture categories was constructed for model training and testing. The smallest trained model, LSP-YOLO-n, achieved 94.2% accuracy and 251 Fps on personal computer(PC) with a model size of only 1.9 MB. Meanwhile, real-time and high-accuracy inference under constrained computational resources was demonstrated on the SV830C + GC030A platform. The proposed approach is characterized by high efficiency, lightweight design and deployability, making it suitable for smart classrooms, rehabilitation, and human-computer interaction applications.

Knowing the reliability of a model's response is essential in application. With the strong generation capabilities of LLMs, research has focused on generating verbalized confidence. This is further enhanced by combining chain-of-thought reasoning, which provides logical and transparent estimation. However, how reasoning strategies affect the estimated confidence is still under-explored. In this work, we demonstrate that predicting a verbalized probability distribution can effectively encourage in-depth reasoning for confidence estimation. Intuitively, it requires an LLM to consider all candidates within the answer space instead of basing on a single guess, and to carefully assign confidence scores to meet the requirements of a distribution. This method shows an advantage across different models and various tasks, regardless of whether the answer space is known. Its advantage is maintained even after reinforcement learning, and further analysis shows its reasoning patterns are aligned with human expectations.

Membership inference attacks (MIAs) on large language models (LLMs) pose significant privacy risks across various stages of model training. Recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have brought a profound paradigm shift in LLM training, particularly for complex reasoning tasks. However, the on-policy nature of RLVR introduces a unique privacy leakage pattern: since training relies on self-generated responses without fixed ground-truth outputs, membership inference must now determine whether a given prompt (independent of any specific response) is used during fine-tuning. This creates a threat where leakage arises not from answer memorization. To audit this novel privacy risk, we propose Divergence-in-Behavior Attack (DIBA), the first membership inference framework specifically designed for RLVR. DIBA shifts the focus from memorization to behavioral change, leveraging measurable shifts in model behavior across two axes: advantage-side improvement (e.g., correctness gain) and logit-side divergence (e.g., policy drift). Through comprehensive evaluations, we demonstrate that DIBA significantly outperforms existing baselines, achieving around 0.8 AUC and an order-of-magnitude higher TPR@0.1%FPR. We validate DIBA's superiority across multiple settings--including in-distribution, cross-dataset, cross-algorithm, black-box scenarios, and extensions to vision-language models. Furthermore, our attack remains robust under moderate defensive measures. To the best of our knowledge, this is the first work to systematically analyze privacy vulnerabilities in RLVR, revealing that even in the absence of explicit supervision, training data exposure can be reliably inferred through behavioral traces.

Abstract--Random Forests (RFs) typically train each tree on a bootstrap sample of the same size as the training set, i.e., bootstrap rate (BR) equals 1.0. We systematically examine how varying BR from 0.2 to 5.0 affects RF performance across 39 heterogeneous regression datasets and 16 RF configurations, evaluating with repeated two-fold cross-validation and mean squared error . Our results demonstrate that tuning the BR can yield significant improvements over the default: the best setup relied on BR 1.0 for 24 datasets, BR > 1.0 for 15, and BR = 1.0 was optimal in 4 cases only. We establish a link between dataset characteristics and the preferred BR: datasets with strong global feature-target relationships favor higher BRs, while those with higher local target variance benefit from lower BRs. T o further investigate this relationship, we conducted experiments on synthetic datasets with controlled noise levels. These experiments reproduce the observed bias-variance trade-off: in low-noise scenarios, higher BRs effectively reduce model bias, whereas in high-noise settings, lower BRs help reduce model variance. Overall, BR is an influential hyperparameter that should be tuned to optimize RF regression models. ANDOM Forest (RF) is an ensemble machine learning (ML) algorithm involving a set of decision trees that collectively make a decision. In classification tasks, each tree votes for a particular class, and the predicted label is determined either by hard voting (majority vote) or soft voting (averaged class probabilities across the trees). In regression tasks, the final prediction is the mean of all individual tree outputs. RFs serve as a robust baseline across a wide range of ML problems, offering an effective balance of predictive accuracy, training speed, and moderate interpretability. While gradient-boosted trees or deep neural networks may outperform them in heavily tuned or domain-specific settings, RF models consistently deliver near-optimal results with minimal tuning, especially on structured, tabular datasets [1], [2].

Despite the parallel challenges that audio and text domains face in evaluating generative model outputs, preference learning remains remarkably underexplored in audio applications. Through a PRISMA-guided systematic review of approximately 500 papers, we find that only 30 (6%) apply preference learning to audio tasks. Our analysis reveals a field in transition: pre-2021 works focused on emotion recognition using traditional ranking methods (rankSVM), while post-2021 studies have pivoted toward generation tasks employing modern RLHF frameworks. We identify three critical patterns: (1) the emergence of multi-dimensional evaluation strategies combining synthetic, automated, and human preferences; (2) inconsistent alignment between traditional metrics (WER, PESQ) and human judgments across different contexts; and (3) convergence on multi-stage training pipelines that combine reward signals. Our findings suggest that while preference learning shows promise for audio, particularly in capturing subjective qualities like naturalness and musicality, the field requires standardized benchmarks, higher-quality datasets, and systematic investigation of how temporal factors unique to audio impact preference learning frameworks.

This paper proposes a new enhanced model architecture to perform classification of lumbar spine degeneration with DICOM images while using a hybrid approach, integrating EfficientNet and VGG19 together with custom-designed components. The proposed model is differentiated from traditional transfer learning methods as it incorporates a Pseudo-Newton Boosting layer along with a Sparsity-Induced Feature Reduction Layer that forms a multi-tiered framework, further improving feature selection and representation. The Pseudo-Newton Boosting layer makes smart variations of feature weights, with more detailed anatomical features, which are mostly left out in a transfer learning setup. In addition, the Sparsity-Induced Layer removes redundancy for learned features, producing lean yet robust representations for pathology in the lumbar spine. This architecture is novel as it overcomes the constraints in the traditional transfer learning approach, especially in the high-dimensional context of medical images, and achieves a significant performance boost, reaching a precision of 0.9, recall of 0.861, F1 score of 0.88, loss of 0.18, and an accuracy of 88.1%, compared to the baseline model, EfficientNet. This work will present the architectures, preprocessing pipeline, and experimental results. The results contribute to the development of automated diagnostic tools for medical images.

Security operations centers (SOCs) face a persistent challenge: efficiently triaging a high volume of user-reported phishing emails while maintaining robust protection against threats. This paper presents the first randomized controlled trial (RCT) evaluating the impact of a domain-specific AI agent - the Microsoft Security Copilot Phishing Triage Agent - on analyst productivity and accuracy. Our results demonstrate that agent-augmented analysts achieved up to 6.5 times as many true positives per analyst minute and a 77% improvement in verdict accuracy compared to a control group. The agent's queue prioritization and verdict explanations were both significant drivers of efficiency. Behavioral analysis revealed that agent-augmented analysts reallocated their attention, spending 53% more time on malicious emails, and were not prone to rubber-stamping the agent's malicious verdicts. These findings offer actionable insights for SOC leaders considering AI adoption, including the potential for agents to fundamentally change the optimal allocation of SOC resources.

Mobile devices increasingly require the parallel execution of several computing tasks offloaded at the wireless edge. Existing communication systems only support parallel transmissions at the bit level, which fundamentally limits the number of tasks that can be concurrently processed. To address this bottleneck, this paper introduces the new concept of Semantic Multiplexing. Our approach shifts stream multiplexing from bits to tasks by merging multiple task-related compressed representations into a single semantic representation. As such, Semantic Multiplexing can multiplex more tasks than the number of physical channels without adding antennas or widening bandwidth by extending the effective degrees of freedom at the semantic layer, without contradicting Shannon capacity rules. We have prototyped Semantic Multiplexing on an experimental testbed with Jetson Orin Nano and millimeter-wave software-defined radios and tested its performance on image classification and sentiment analysis while comparing to several existing baselines in semantic communications. Our experiments demonstrate that Semantic Multiplexing allows jointly processing multiple tasks at the semantic level while maintaining sufficient task accuracy. For example, image classification accuracy drops by less than 4% when increasing from 2 to 8 the number of tasks multiplexed over a 4$\times$4 channel. Semantic Multiplexing reduces latency, energy consumption, and communication load respectively by up to 8$\times$, 25$\times$, and 54$\times$ compared to the baselines while keeping comparable performance. We pledge to publicly share the complete software codebase and the collected datasets for reproducibility.

Deep ensembles (DE) have emerged as a powerful approach for quantifying predictive uncertainty and distinguishing its aleatoric and epistemic components, thereby enhancing model robustness and reliability. However, their high computational and memory costs during inference pose significant challenges for wide practical deployment. To overcome this issue, we propose credal ensemble distillation (CED), a novel framework that compresses a DE into a single model, CREDIT, for classification tasks. Instead of a single softmax probability distribution, CREDIT predicts class-wise probability intervals that define a credal set, a convex set of probability distributions, for uncertainty quantification. Empirical results on out-of-distribution detection benchmarks demonstrate that CED achieves superior or comparable uncertainty estimation compared to several existing baselines, while substantially reducing inference overhead compared to DE.

To mitigate the potential harms of Large Language Models (LLMs)generated text, researchers have proposed watermarking, a process of embedding detectable signals within text. With watermarking, we can always accurately detect LLM-generated texts. However, recent findings suggest that these techniques often negatively affect the quality of the generated texts, and adversarial attacks can strip the watermarking signals, causing the texts to possibly evade detection. These findings have created resistance in the wide adoption of watermarking by LLM creators. Finally, to encourage adoption, we evaluate the robustness of several watermarking techniques to adversarial attacks by comparing paraphrasing and back translation (i.e., English $\to$ another language $\to$ English) attacks; and their ability to preserve quality and writing style of the unwatermarked texts by using linguistic metrics to capture quality and writing style of texts. Our results suggest that these watermarking techniques preserve semantics, deviate from the writing style of the unwatermarked texts, and are susceptible to adversarial attacks, especially for the back translation attack.