Agent-based neural models often encode transmission within neuron state updates, which can make it difficult to separately log and quantify spatial recruitment patterns, delay structure, and environment-mediated feedback effects. We present LANA (Local Adaptive Neural Agents), a dual-agent neural agent-based model in which neurons and propagating signals are represented as distinct interacting entities embedded in a dynamic environmental field. The model combines discrete leaky integrate-and-fire neuron dynamics, mobile signal agents, synaptic links with distance-dependent delays, and a bounded environment-to-neuron feedback mechanism. LANA is intended as a normalized phenomenological mesoscopic framework for mechanism-level comparison rather than as a circuit-specific biophysical reconstruction. To support interpretability and reproducibility, we report a compact internal verification block for the implemented operators, including delay propagation, environmental decay and diffusion, threshold activation, and refractory enforcement. We then compare the full LANA model against a matched neuron-only baseline and summarize spatial recruitment using first-spike maps, cumulative recruitment times, and wavefront speed as a secondary descriptive metric. Finally, we evaluate two controlled operating regimes, a resting regime (S1) and a hyperexcitable regime (S2), under fixed network size, stimulation schedule, and matched random seeds. Relative to the baseline, the full model sustains and spreads activity more effectively and provides spatially resolved recruitment summaries, including first-spike timing and cumulative recruitment measures, that are not available in the same form when transmission is represented only through neuron-level updates. Relative to S1, S2 exhibits earlier activation, higher firing activity, stronger environmental accumulation, and faster cumulative recruitment. Local and factorial sensitivity analyses further identify the parameters that most strongly govern these regime differences. Together, these results position LANA as a normalized mesoscopic and computationally tractable framework for studying how excitability, transport state dynamics, delayed coupling, and environment-mediated feedback jointly shape emergent activity in controlled simulation settings.

Spiking neural networks can exhibit complex emergent dynamics, but the credibility of spatially explicit agent-based implementations depends on systematic verification and validation (V&V). This study introduces LANA (Local Adaptive Neural Agents), an agent-based spiking neural network in which neurons, propagating signals, directed synapses, and a diffusive environmental field are represented as distinct interacting components. We present a five-level V&V framework spanning operator-level tests, single-neuron mechanisms, propagation behavior, network-level dynamics, and sensitivity/robustness analysis. Across 13 predefined tests and approximately 2000 simulation runs, the model satisfied all prespecified pass criteria: synaptic delays reproduced the expected propagation law exactly, environmental decay and diffusion matched analytical expectations, threshold and refractory mechanisms behaved as predicted, inhibition suppressed firing monotonically, and environmental coupling induced a transition toward higher variability and oscillatory-like activity. Matched-seed comparisons further showed that explicit signal transport and environmental feedback substantially amplify activity relative to a neuron-only baseline while leaving synaptic delay propagation unchanged. Additional regime and lesion experiments demonstrated distinct resting, hyperexcitable, and focal-lesion states, with the lesion condition producing an acute decline followed by only partial recovery. Together, these results provide a transparent V&V baseline for LANA and illustrate how agent-based spiking models can be tested and interpreted across multiple scales.

Selecting a machine learning model for higher-education quality assurance is a multi-criteria decision problem that cannot be reduced to a single leaderboard metric. This paper presents a two-step disclosure protocol in which users first make an unaided model choice and then revise it after structured multi-criteria disclosure including criterion weights, normalized values, and a ranked recommendation. The overall study used a three-step experimental protocol, with the disclosure manipulation itself implemented as a two-stage intervention within Step 2. The protocol was evaluated with 38 participants under a stable Dean-oriented advisory framing across three institutional prediction tasks, yielding 228 confirmatory scenarios after predefined quality filters. Decision quality was operationalized as regret reduction relative to a frozen governance-oriented multi-criteria scoring policy. Results show that structured disclosure significantly improved policy-aligned decision quality (Wilcoxon p = 2.30 × 10⁻11, rank-biserial r = 0.864) and increased self-reported decision confidence (p = 7.62 × 10⁻12, r = 0.646). Importantly, significant improvement was already observed in the information-only stage before any explicit recommendation was shown (p = 5.18 × 10⁻5, r = 0.629). Post-decision trust changes were small and did not reach significance in the confirmatory analysis, and are therefore treated as exploratory. The findings provide protocol-level evidence that structured multi-criteria disclosure can improve alignment with a predefined governance-oriented model selection policy in educational QA settings.

Outcome improvement alone does not reveal whether users actually inspected disclosed evidence or simply followed a highlighted recommendation. This companion human-study paper analyzes model-selection deliberation under a staged multi-criteria disclosure interface for educational quality assurance (QA). In the final filtered analytic sample of 38 participants and 228 completed scenarios, we examine interface telemetry, participant-level self-report, acceptance, task-level heterogeneity, and a heuristic low-engagement robustness check. Nonparametric comparisons and a clustering-adjusted GEE model were used. Disclosure uptake was selective: ranking was used in 60.5% of scenarios, weights in 47.4%, heatmap in 46.9%, and textual interpretations in 41.7%. Self-reports aligned with telemetry for four of the five major components. Improved scenarios showed longer Step 2 deliberation and greater engagement with multiple evidence surfaces, while the GEE model indicated that ranking use was significantly associated with improvement. Acceptance was favorable (34/38 preferred the agent-assisted workflow), whereas a low-engagement subgroup showed sharply reduced benefit. The observed pattern is more consistent with structured multi-surface deliberation than with shallow recommendation following.

Implementing a cognitive Sense-Think-Act-Learn (STAL) architecture for automated cultural heritage visualization, this paper illustrates a multi-agent system. The system consists of three specialized agents: an Exhibition Curator Agent that applies Large Language Models (LLMs) to classify and generate exhibitions, a Conversational Guide Agent that provides interactive visitor engagement, and an Image Acquisition Agent that performs automated visual content enrichment. Evaluated on a dataset including 6,398 historical events across multiple civilizations and epochs, the system effectively automates the curation of the data by adopting a hybrid approach via deterministic classification algorithms as well as LLM-based analysis. The architecture allows adaptive learning based on administrative feedback, improving classification accuracy over time. This work provides a concrete framework to leverage generative $A I$ in cultural heritage digitization while also addressing issues related to scale, multilingual content, and domain-specific curation needs.

This paper presents the LANA Adaptive Labeling Framework (ALF) as an advanced framework for dynamic method labeling and selecting optimal data processing methods in multiple multicriteria intelligent software systems, focusing on business processes in higher education institutions (HEIs). Earlier approaches to method labeling relied on static hierarchical structures. In contrast, LANA ALF introduces adaptability through continuous learning from user feedback, automatic balancing of criteria based on historical data and current task requirements, and multidimensional labels for comprehensive method evaluation. Each query is represented with a set of labels, while neural networks evaluate the optimal method by balancing criteria such as performance, cost, reliability, and accuracy. User feedback is stored in dynamic tables (e.g., user satisfaction), automatically adapting their structure to new tasks and data types. The results demonstrate that LANA ALF enables intelligent agents to autonomously make decisions without the need for direct involvement of data science experts, thereby increasing accuracy, reliability, and user satisfaction. This framework provides a foundation for further application of ALF in various domains