Davis & Elkins College (D&E) is a small private liberal arts college in Elkins, West Virginia..
Growing evidence suggests that layer attention mechanisms, which enhance interaction among layers in deep neural networks, have significantly advanced network architectures. However, existing layer attention methods suffer from redundancy, as attention weights learned by adjacent layers often become highly similar. This redundancy causes multiple layers to extract nearly identical features, reducing the model's representational capacity and increasing training time. To address this issue, we propose a novel approach to quantify redundancy by leveraging the Kullback-Leibler (KL) divergence between adjacent layers. Additionally, we introduce an Enhanced Beta Quantile Mapping (EBQM) method that accurately identifies and skips redundant layers, thereby maintaining model stability. Our proposed Efficient Layer Attention (ELA) architecture, improves both training efficiency and overall performance, achieving a 30
Large Vision-Language Models (LVLMs) have achieved remarkable progress in visual understanding tasks such as image captioning and visual question answering. However, they remain susceptible to hallucinations, generating content that is inconsistent with the actual visual input. Existing methods primarily intervene at the decoding stage, while overlooking a critical source of hallucinations: irrelevant or noisy visual tokens that mislead the decoding process. To address this issue, we propose SeeMe, a training-free framework that introduces the concept of feature engineering from traditional machine learning into LVLMs. SeeMe restructures visual tokens through a three-stage token engineering process to suppress hallucination sources while preserving informative visual evidence. Experiments on MME, POPE, and AMBER benchmarks across four LVLMs demonstrate that SeeMe consistently reduces hallucinations and improves output consistency, providing a novel perspective for mitigating hallucinations in LVLMs.
Large vision-language models (LVLMs) achieve strong multimodal understanding, but the hundreds to thousands of visual tokens they process impose substantial computational overhead, motivating training-free visual token pruning. In this work, we conduct two complementary analyses of visual token pruning. First, we measure the feature-space coverage of tokens retained before cross-modal fusion and find that aggressive pruning discards substantial visual information. Second, we track text-to-visual attention across decoder layers and find that the visual tokens considered important change substantially with depth, making one-shot pruning decisions unreliable. Together, these findings show that effective pruning should preserve broad visual coverage before fusion and progressively refine the retained tokens as cross-modal evidence evolves during fusion. We therefore propose STAR-Pro (STage-Wise Adaptive Token Reduction with Progressive Refinement), a training-free two-stage framework. Its Adaptive Stage applies pivoted QR to construct an over-budget feature-coverage candidate pool, while its Progressive Stage uses evolving text-to-visual attention at selected decoder layers to prune a nested survivor set under a target layer-average token budget. Extensive experiments across seven LVLMs spanning multiple architectures and 18 image and video benchmarks demonstrate the effectiveness of STAR-Pro under aggressive pruning. On LLaVA-Video-7B, STAR-Pro reduces visual tokens by 90.5
Organizations are, in hindsight, facing an economic crisis due to quiet quitting. This is a quantitative cross-sectional study on the antecedents of quiet quitting: perceived organizational justice (distributive, procedural, interpersonal, and informational), work engagement, and organizational commitment (affective, normative, and continuance). A total of 229 samples were collected from five private primary and secondary educational institutions in the Philippines, of which 180 qualified for data analysis. Data were collected from employees with 12 months of service in their institution. Multiple linear regression using JASP was utilized for data analysis. The results show that low work engagement, low affective commitment, and high continuance commitment are the predictors most likely to predict quiet quitting, whereas perceived distributive, procedural, interpersonal, and informational justice, as well as normative commitment, are not. This study contributes to new knowledge on the antecedents of quiet quitting in the context of education. The results of this study serve as input for policy development and interventions of school administrators and human resource practitioners.
de Broglie-Bohm theory (dBBT), treating quantum particles as point objects moving along well defined (Bohmian) trajectories, offers an appealing solution of the measurement problem in quantum mechanics; it has, however, problems relating to spin, relativity and lack of proper integration with the Hilbert space based framework. In this work, we present a consistent Hilbert space based formalism which has the traditional state-observable framework integrated with the desirable features of dBBT. We adopt ensemble interpretation for the Schr ö dinger wave function ψ . Using a fixed time wave function ψ _0 to make the system configuration space M (= ℝ^n) a probability space ℳ_0 = (ℝ^n, |ψ _0|^2 dx) , we introduce a stochastic process ξ (t) such that its expectation value in ℳ_0 equals that of the Heisenberg position operator X_H(t) in the Heisenberg state |ψ _H⟩ corresponding to ψ _0 . This condition leads to the de Broglie-Bohm guidance equation for the sample paths of the process ξ (t) which are, therefore, Bohmian trajectories supposedly representing time-evolutions of individual members of the ψ _0 -ensemble. Stochastic processes and Bohmian trajectories corresponding to observables with discrete eigenvalues (in particular spin) are treated by extending the configuration space to the spectral space of the commutative algebra obtained by adding appropriate discrete observables to the position observables. Pauli’s equation (the Schr ö dinger equation for a nonrelativistic charged spin half particle) is treated as an example. A straightforward derivation of von Neumann’s projection rule employing the Schr ö dinger - Bohm evolution of individual systems along their Bohmian trajectories is given. Some comments on the potential application of the formalism developed here to quantum mechanics of the universe are included.
