End-to-end training of multimodal neural networks often exhibits unstable neural dynamics characterized by three coupled failure modes that degrade learning: (i) modality imbalance, where one branch dominates gradient-based optimization; (ii) unstable gating, where noisy confidence cues induce erratic modality selection; and (iii) fusion interference, where modality-specific gradients conflict at the shared fusion layer. We propose CAT-GS (Calibrated, Adaptive, Thresholded Gating with Fusion Surgery), a neural dynamics-based optimization controller for intelligent computing applications. CAT-GS operates during backpropagation without modifying model architectures, fusion modules, or task losses. Through calibration of teacher-derived reliability via temperature scaling and EMA smoothing, CAT-GS stabilizes neural dynamics using a margin-thresholded policy to switch between warm-up dropout, weak-modality prioritization, and weak-biased blending, stabilizes gradient magnitudes under aggressive gating via capped gradient-budget renormalization, and applies fusion-only PCGrad to reduce destructive cross-modal interference at the primary shared bottleneck. We evaluate CAT-GS on audio–visual multimodal pattern recognition benchmarks (CREMA-D, AV-MNIST, and VGGSound), a tri-modal setting (UR-FUNNY), controlled synthetic data (CG-MNIST), and additional cross-domain benchmarks (AVE and CMU-MOSI). CAT-GS improves or matches fused multimodal accuracy against strong imbalance-aware baselines (including OGM-GE, G^2D, and UMT) across settings, and yields smoother gating behavior with fewer conflicting fusion gradients.
This study links mobility styles with carsharing intentions and next-generation vehicle preferences in urban Japan. An online survey of licensed drivers in Tokyo Metropolis, Kanagawa Prefecture, and Osaka Prefecture captured mode use frequencies and stated preferences by travel purpose. Cluster analysis identified three mobility styles: private-car users, public-transport users, and a diversified, high-carsharing group. The diversified profile combined high carsharing participation with high private-car ownership, indicating that carsharing can complement rather than replace existing car access. Random Forest results indicate that age, car usage frequency, and income best distinguish carsharing users from non-users. Carsharing preferences varied substantially by stated-use context. For daily and business trips, most respondents prefer carsharing stations within 400 m and short distances; for tourism and hometown visiting, they accept longer access to carsharing stations and longer trips. Compact and Kei cars (Japanese light vehicles) in Japan dominate daily use, while minivans and SUVs gain importance for leisure. Gasoline and hybrid vehicles are most preferred, with limited willingness to pay extra for electric vehicles. The findings support segment-and purpose-specific carsharing station placement, fleet composition, and gradual electrification strategies in mature, transit-rich metropolitan areas.
As the Paris Agreement completes its 10th year in effect, it becomes duty to reflect on the impact it has had thus far. While many praises are to be sung for our well-sung hero, the agreement still leaves much to be desired, and consequently has been subject to harsh criticism. Yet in the growing body of literature inspecting the agreement with a scrutinous eye, few studies seem to adequately question the underlying assumption of sustained cooperation under the Paris Agreement, especially through the lens of ‘energy constraints’. Even research which assesses cooperation in international environmental agreements (IEAs) seems to largely ignore this fundamental limitation. The paper adopts a critical review approach to address this gap and, at the same time, develops a new conceptual framework to assess the Paris Agreement as an “energy constraint two-level repeated game”, incorporating biophysical economics with the study of cooperation in IEAs. As per our proposed framework, binding energy constraints can potentially undermine cooperation under the Paris Agreement, as inability to adequately meet domestic energy needs can trigger downstream effects on political will to continue cooperation under the agreement. As such, the Paris Agreement, or any other agreement must first ensure compatibility with energy constraints to make the notion of cooperation credible.
Since 1978, the Rohingya population has faced multiple humanitarian crises, including racial disparities, which significantly escalated during 2017, leading to widespread exile from Myanmar. Reportedly, violence, persecution, and trauma have posed grievous impacts on the mental health of these forcibly displaced people. This scoping review aimed to synthesize the evidence regarding the overall epidemiologic burden of psychological problems of Rohingya refugees with their associated factors. We evaluated five major databases and additional sources till July 31, 2024, and included articles according to the eligibility criteria following the Joanna Briggs Institute (JBI) guidelines and the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist. Out of 373 citations retrieved from multiple sources, we included 35 articles in this review. Most of the articles reported a high prevalence of different psychological symptoms of Rohingya refugees, such as depression, anxiety, post-traumatic stress disorder, persistent complex bereavement disorder, feeling afraid, etc. Several correlates of mental health problems were reported, including older age, being female, illiteracy, experiences of torture, sexual violence, unemployment, food insecurity, statelessness, lack of healthcare access, unhygienic campsites, and preexisting health problems, etc. There were significant gaps in community-level intervention studies, however, Group Integrated Adapt Therapy (IAT-G) and Mental Health and Psychosocial Support (MHPSS) services are widely used. The available evidence suggests a huge burden of mental health disorders with several biopsychosocial factors that may assist in better policymaking and implementation of multilayered approaches, like improving healthcare access, training healthcare providers, more community-based intervention studies, and introducing tele-mental health services for Rohingya refugees.
Magnetically actuated nanobeams have emerged as important structural elements in nanoelectromechanical systems because longitudinal magnetic fields enable contactless tuning of resonance characteristics. Since nanoscale beam dynamics are strongly influenced by size effects that cannot be captured adequately by classical local theories, physically consistent size-dependent formulations are required for reliable vibration prediction. In this study, a stress-driven nonlocal Rayleigh beam model is developed to investigate the free vibration behavior of conducting nanobeams subjected to a longitudinal magnetic field. Closed-form analytical solutions are derived for different boundary conditions, allowing the coupled effects of nonlocality, magnetic interaction, and inertia to be examined in a unified and physically consistent manner. The results reveal that the application of a longitudinal magnetic field induces a systematic stiffening effect, leading to an increase in natural frequencies for all boundary conditions and vibration modes considered. The degree of magnetic sensitivity is shown to depend strongly on the support configuration, with simply supported nanobeams exhibiting higher relative frequency tuning compared to clamped counterparts. The influence of stress-driven nonlocality is found to modify both the baseline frequency spectrum and the magnetic stiffening trends, particularly for higher modes where rotary inertia plays a significant role. Distinct from earlier strain-driven nonlocal magnetic analyses, the present stress-driven formulation avoids boundary-condition inconsistencies and yields physically meaningful predictions across all cases examined. The proposed model and benchmark results are expected to be useful for validating numerical simulations and for providing theoretical guidance in the analysis of magnetically tunable nanoscale resonators and related devices.
