We propose JEPADepth, a self-supervised monocular depth estimation framework that augments photometric view synthesis with an auxiliary masked-prediction loss computed in the representation space of a pretrained DINOv3 Vision Transformer (ViT), following the Image Joint-Embedding Predictive Architecture (I-JEPA) principle. A predictor infers target-region embeddings from visible context-region embeddings under structured masking; the predictor and target encoder are discarded at inference, so the deployed 24 M-parameter model costs no more than the photometric-only baseline. On the KITTI Eigen split, adding the JEPA objective to an identical DINOv3 photometric baseline lowers absolute relative error from 0.105 to 0.101 and raises δ1 from 0.891 to 0.898, a gain roughly four times the run-to-run standard deviation measured over five random seeds. On the improved ground truth, it reaches 0.075 absolute relative error, matching the best transformer competitor while using 11% fewer inference parameters. Under zero-shot transfer without fine-tuning, it obtains the best absolute relative error on Make3D (0.275) and ties the best on Cityscapes (0.143) while leading on five of seven metrics. Training overhead is ∼16%; inference cost is unchanged.
Regulatory instruments alter institutional payoffs, but they do not choose which equilibrium a decentralized financial system will realize. A policy that appears safe at one computed response can therefore coexist with an unobserved fire-sale or funding-run outcome. We pose policy construction as a set-containment requirement: every approximate Nash equilibrium of the regulated game must remain inside a buffered stability target. EquiImplement enforces this requirement with state-dependent mixtures of admissible unilateral deviations over the target complement. When policy features enter payoffs affinely, holding this proof object fixed gives affine inequalities in the regulatory vector and a concave minimum margin, producing a convex semi-infinite policy subproblem. A rigorous regional separator locates the weakest-covered state, and the same verified counterexample drives both the next policy solve and certificate revision. A preset outer-iteration budget gives finite execution and one of three outputs: globally certified containment, budget infeasibility within the prescribed incentive–certificate family, or an unresolved status. The evaluation covers 330 fire-sale and rollover/run games. At the default budget, EquiImplement attains full-stability rates of 95.0%, 85.6%, and 92.5%, while reducing the worst-equilibrium violation from 0.037 for the strongest non-oracle baseline to 0.016. The mean regulatory cost over returned budget-feasible evaluation policies decreases from 0.119 to 0.113. Further analyses relate certification difficulty to stability-target tightness, leverage, market depth, liquidity buffers, and coordination strength. The results show that a deviation-based proof object can serve directly as an optimization interface for financial regulation under equilibrium multiplicity.
Efficient subsidy allocation for heterogeneous community-based elderly care services requires identifying latent service demand and optimizing service providers’ responses under fiscal constraints. To address this problem, this study develops a data-driven Stackelberg game framework for demand-aware subsidy optimization. The framework transforms unstructured user-generated reviews into empirical demand-priority weights through Latent Dirichlet Allocation (LDA) topic modeling and incorporates these data-derived parameters into a budget-constrained Stackelberg game to optimize differentiated subsidy decisions. A total of 11,257 elderly care-related reviews from e-commerce platforms were collected and refined into 3907 valid samples. Six major service needs were identified: Daily Living Care, Meals and Nutrition, Smart Safety Monitoring, Medical Assistance, Professional Nursing, and Home Rehabilitation. These demand weights were incorporated into the optimization model to characterize heterogeneous service priorities and determine equilibrium subsidy decisions. The numerical experiments show that differentiated subsidies substantially improve demand-weighted service coverage compared with demand-unaware allocation under the same fiscal constraint. At a budget of RMB 300,000, the demand-aware model achieves a demand-weighted coverage of 0.8328, compared with 0.7309 for the demand-unaware model. Budget sensitivity analysis further shows diminishing marginal returns to fiscal expansion, with full potential demand reached at the model-based threshold of RMB 479,500. Robustness analysis indicates that the main allocation results remain stable under perturbations of the empirically derived demand-priority weights.
Existing elevator-dispatch research uses quadratic waiting-cost models and restricts reassignment near pickup to preserve service decisions. Although fixing a hall-call assignment removes it from the decision variables, its service remains on the route, so omitting its interactions changes the objective for the remaining calls. This study examines the resulting omission loss and the computation of the retained interactions. Conditioning on fixed assignments yields a unary correction. Its omission loss is bounded, and reachable three-floor states approach the bound even for arbitrarily short reassignment thresholds. Under explicit service assumptions, floor-ordered pair-cost formulas include the merged-turnaround boundary and permit linear arithmetic for supported fixed-call interaction sums after unary evaluation. Across 30 paired seeds and 75 synthetic interfloor settings, adding the correction to move-and-swap search reduced mean waiting time by 0.50 s (95% interval 0.40–0.59 s). Separately, structured coefficient assembly reduced the equal-cell mean of complete dispatch time from 46.039 to 36.945 microseconds on identical inputs, a 19.8% reduction. The omission bound and interaction formulas provide an analytical basis for retaining fixed-call service costs, with measured service and computational benefits in the tested interfloor regime.
This paper presents a partial differential equation (PDE)-based auto-aggregation model for simulating descriptive norm dynamics in autonomous multi-agent systems, capturing convergence and violation through nonlocal perception kernels and external potential fields. Extending classical transport equations, the framework represents opinion popularity as a continuous distribution, enabling direct interactions without Bayesian guessing of beliefs. Using azithromycin and prednisolone as illustrative treatment features from a real-world coronavirus disease 2019 (COVID-19) dataset, the numerical experiments exhibit reduced local perception-to-target discrepancy under target-based updating, redistribution and the formation of a new unimodal density under fact-informed updates with potential guidance, and persistent multiple peaks without target-based fitting or an external potential field. Quantitative evaluation uses population-weighted local perception-to-target Wasserstein distance estimates for the selected treatment dimensions.
