Perovskite solar cells (PSCs) have emerged as promising photovoltaic technologies owing to their high power conversion efficiency (PCE) and material versatility. Conventional optimization of PSC architectures largely depends on iterative experimental approaches, which are often labor-intensive and time-consuming. In this study, a data-driven modeling strategy is introduced to accelerate the design of efficient and mechanically robust PSCs. Seven supervised regression models were evaluated for predicting key photovoltaic parameters, including PCE, short-circuit current density (Jsc), open-circuit voltage (Voc), and fill factor (FF). Among these, a stacking ensemble framework exhibited superior predictive accuracy, achieving an R² of 0.8577 and a root mean square error of 2.084 for PCE prediction. Model interpretability was ensured through Shapley Additive exPlanations(SHAP) analysis, which identified precursor solvent composition, A-site cation ratio, and hole-transport-layer additives as the most influential parameters. Guided by these insights, ten device configurations were fabricated, achieving a maximum PCE of 24.9%, in close agreement with model forecasts. Furthermore, multiscale mechanical assessments, including bending, compression, impact resistance, peeling adhesion, and nanoindentation tests, were conducted to evaluate structural reliability. The optimized device demonstrated enhanced interfacial stability and fracture resistance, validating the proposed predictive–experimental framework. This work establishes a comprehensive approach for performance-oriented and reliability-driven PSC design, providing a foundation for scalable and durable photovoltaic technologies.

Large Language Models (LLMs) excel in diverse text generation tasks but still face limited controllability, opaque decision processes, and frequent hallucinations. This paper presents a structural causal intervention framework that models input–hidden–output dependencies through a structural causal model and performs targeted interventions on hidden representations. By combining counterfactual sample construction with contrastive training, our method enables precise control of style, sentiment, and factual consistency while providing explicit causal explanations for output changes. Experiments on three representative tasks demonstrate consistent and substantial improvements: style transfer accuracy reaches 92.3% (+7–14 percentage points over strong baselines), sentiment-controlled generation achieves 90.1% accuracy (+1.3–10.9 points), and multi-attribute conflict rates drop to 3.7% (a 40–60% relative reduction). Our method also improves causal attribution scores to 0.83–0.85 and human agreement rates to 87–88%, while reducing training and inference latency by 25–30% through sparse masking that modifies ≤10% of hidden units per attribute. These results confirm that integrating structural causal intervention with counterfactual training advances controllability, interpretability, and efficiency in LLM-based generation, offering a robust foundation for deployment in reliability-critical and resource-constrained applications.

Conventional teacher assessment is labor-intensive and subjective. Prior LLM-based systems improve scale but rely on post hoc rationales and lack built-in trust controls. We propose an explainable-by-design framework that couples (i) Dual-Lens Hierarchical Attention—a global lens aligned to curriculum standards and a local lens aligned to subject-specific rubrics—with (ii) a Trust-Gated Inference module that combines Monte-Carlo-dropout calibration and adversarial debiasing, and (iii) an On-the-Spot Explanation generator that shares the same fused representation and predicted score used for decision making. Thus, explanations are decision-consistent and curriculum-anchored rather than retrofitted. On TeacherEval-2023, EdNet-Math, and MM-TBA, our model attains an Inter-Rater Consistency of 82.4%, Explanation Credibility of 0.78, Fairness Gap of 1.8%, and Expected Calibration Error of 0.032. Faithfulness is verified via attention-to-rubric alignment (78%) and counterfactual deletion tests, while trust gating reduces confidently wrong outputs and triggers reject-and-refer when uncertainty is high. The system retains 99.6% accuracy under cross-domain transfer and degrades only 4.1% with 15% ASR noise, reducing human review workload by 41%. This establishes a reproducible path to trustworthy and pedagogy-aligned LLMs for high-stakes educational evaluation.