generative ai

This work addresses key bottlenecks in high-dimensional supervised learning and Bayesian inference—namely, strong parametric assumptions and reliance on large-scale, accurately labeled real-world data. We propose a model-agnostic generative modeling framework that leverages generative AI (Gen-AI) to synthesize high-fidelity training samples and employs deep neural networks for end-to-end nonparametric estimation of conditional densities and posterior quantiles—without assuming any prespecified distributional form. The framework unifies high-dimensional regression, dimensionality reduction (including feature selection), and uncertainty quantification. Experiments on the Ebola epidemic dataset demonstrate substantial improvements over conventional parametric methods in predictive accuracy, calibration, and computational scalability. To our knowledge, this is the first scalable, plug-and-play paradigm for model-free density estimation and Bayesian inference.

This work proposes a unified framework for understanding and developing generative artificial intelligence models capable of producing multimodal content, including images, text, video, and molecular structures. Addressing the current fragmentation in generative modeling, the study integrates core methodologies—such as variational autoencoders, generative adversarial networks, diffusion models, and large language models—into a cohesive theoretical system grounded in mathematical principles, architectural design, and mechanisms for controllable generation. This framework not only advances a systematic understanding of multimodal generative processes but also provides robust theoretical foundations and practical pathways for generating high-quality, controllable digital content, with direct implications for applications in scientific discovery and beyond.

This paper addresses a foundational question in generative AI: What is its intrinsic nature as a distinct machine learning task, and how can generative tasks be formally characterized and theoretically related to prediction, compression, and decision-making? To this end, the authors propose a task-centric research paradigm and develop a unified theoretical framework integrating probabilistic modeling and two-player game theory, rigorously distinguishing density estimation from sampling-based generation. The framework systematically unifies five major generative paradigms—autoregressive models, variational autoencoders, normalizing flows, generative adversarial networks, and diffusion models—and incorporates post-training alignment strategies while embedding socio-ethical considerations. The resulting formal foundation advances the theoretical understanding of generative AI and provides systematic support for responsible AI practices, including privacy-preserving generation, content provenance, and copyright-compliant deployment.

Generative AI inherently lacks guarantees of correctness, safety, and fairness due to its reliance on probabilistic sampling, necessitating urgent improvements in reliability, evaluation quality, and experimental rigor. This project establishes a statistically grounded theoretical framework for trustworthy generative AI, proposing a principled enhancement pathway integrating probabilistic modeling, Bayesian inference, robust sampling, and causal experimental design—enabling uncertainty quantification and attribute-constrained generation. Innovatively, it unifies statistical inference with AI evaluation, markedly improving evaluation efficiency and discriminative power; it further develops intervention-based experimental paradigms tailored to generative models, supporting causal reasoning for reliability validation. The work clarifies statistics’ central role in generative AI trustworthiness, delivering both a methodological foundation and practical guidelines for building verifiable, controllable, and accountable next-generation AI systems.

This study systematically investigates the security challenges arising as generative AI transitions from content generation to performing real-world actions, introducing a novel tripartite threat taxonomy encompassing content-level, model-level, and agent-level risks. Through integrated threat modeling, evaluation of technical countermeasures—including detection, watermarking, alignment techniques, and agent-specific safeguards—and analysis of governance structures, the work reveals a pervasive gap between the rapid expansion of attack surfaces and the current state of defensive capabilities. Most existing technical solutions remain contingent on nascent institutional coordination mechanisms that have yet to mature. The research underscores the necessity for parallel evolution of technical and governance approaches and highlights the critical importance of cross-layer collaborative defense strategies to effectively mitigate emerging threats.

This paper addresses the conceptual and methodological challenges in bridging machine learning and computational creativity. It systematically traces the evolution of computational creativity theory and surveys key generative deep learning techniques—namely, Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), and Transformers—alongside their applications in creative tasks. To overcome persistent evaluation bottlenecks, the authors propose a hybrid assessment framework integrating cognitive modeling, multi-dimensional aesthetic metrics, and human-grounded benchmarks—marking the first comprehensive integration of theoretical paradigms with generative model practice. The core contributions are threefold: (1) construction of the most comprehensive research map of computational creativity to date; (2) a cross-paradigmatic methodological reflection on evaluation; and (3) identification of explainability enhancement and human-AI co-creation as critical frontiers—thereby establishing theoretical consensus and practical guidelines for algorithm design, evaluation standardization, and interdisciplinary deployment.

This study addresses the widespread lack of understanding of generative AI among energy sector employees, which hinders the identification of viable application entry points and implementation pathways. Through semi-structured interviews, internal document analysis, and on-site observations, the research systematically identifies five high-potential application scenarios: report generation, forecasting, data processing, equipment maintenance, and anomaly detection. It proposes a gradual deployment approach for generative AI that aligns with existing workflows. This work establishes the first practical framework for implementing generative AI in the energy industry, offering reusable design strategies for LLM-based agent workflows and providing both theoretical grounding and actionable guidance for AI-driven transformation in heavy industrial sectors.

This study addresses low development efficiency and poor quality assurance in healthcare Web systems. It pioneers the systematic integration of generative AI into clinical research–oriented medical software, spanning the entire software engineering lifecycle—project management, requirements analysis, system design, coding, and testing. By embedding large language models (LLMs) into established software engineering practices, the approach enables automated requirements documentation, AI-assisted architectural design, context-aware code snippet recommendation, and intelligent test case generation, complemented by a human-in-the-loop quality verification mechanism. Empirical evaluation demonstrates a 35% average reduction in documentation and coding time, improved requirements consistency, and enhanced test coverage. The work yields a reusable, AI-augmented medical software engineering framework and validated best-practice guidelines, providing both theoretical foundations and empirical evidence for the scalable, compliant deployment of generative AI in highly regulated healthcare domains.

Despite the widespread industrial adoption of generative AI, higher education—particularly in computer science—lags significantly, emphasizing theoretical foundations while neglecting hands-on tool proficiency, thereby leaving students unprepared for responsible, professional AI application. Method: This study designed and delivered a generative AI applications course for undergraduate and master’s students, structured modularly around authentic development tasks—including code generation and document automation—and integrating ethical reflection and empirical evaluation. Contribution/Results: As the first systematic integration of generative AI tool competency into the core CS curriculum in China, the course demonstrably enhanced students’ conceptual understanding and practical efficacy, as validated by mixed-methods evaluation. It establishes a scalable, transferable pedagogical framework that effectively bridges the industry–academia skills gap in AI literacy and application.

This study addresses the limitations of general-purpose large language models (LLMs) in specialized domains such as transportation engineering, where insufficient understanding of technical standards and domain-specific semantics hinders performance on complex tasks. To overcome this, the authors construct a domain-specific corpus comprising U.S. transportation engineering manuals, design guidelines, and regulatory documents, and apply continual pretraining to six state-of-the-art LLMs within a unified low-rank adaptation (LoRA) framework. This work introduces the first reproducible paradigm for developing domain-tailored generative AI agents. Experimental results demonstrate that Qwen2.5-7B and LLaMA-3.1-8B achieve superior domain alignment and response quality, confirming that LoRA-based fine-tuning with authoritative technical documentation significantly enhances model comprehension and reasoning capabilities in specialized engineering contexts.

Large-scale Cyber-Physical Systems (CPS) simulation and testing face prohibitively high resource consumption and maintenance costs in modeling hardware, software, and physical environments. Method: Through cross-enterprise workshops involving six industrial partners, this study systematically identifies key engineering bottlenecks and proposes three priority research directions: (1) AI-driven generation of dynamic scenario and environmental models; (2) co-integration of simulators and generative AI into CI/CD pipelines; and (3) trustworthiness assurance mechanisms for generative AI outputs in simulation contexts. Contribution/Results: The work delivers empirically grounded, forward-looking outcomes—distilling a reusable engineer challenge taxonomy and establishing an industry–academia collaborative research agenda. It provides both theoretical foundations and practical guidelines for the responsible deployment of generative AI in CPS simulation and testing.