psychological theory integration

Psychology lacks a mathematically rigorous, interdisciplinary-accessible framework capable of supporting cognitive modeling in psychosomatic medicine and AI safety. This paper introduces a diagrammatic category-theoretic approach grounded in process theory, formally encoding core psychodynamic constructs—such as conflict, defense, and integration—as computable graph structures, thereby yielding an analyzable, dynamical systems model of mental processes. The framework unifies representations of psychological processes, neurobiological mechanisms, and agent-level behavioral logic, establishing formal interfaces between psychology and AI alignment, autonomous agent negotiation, and neuromodulatory intervention. Empirical evaluation demonstrates its utility in individualized AI cognitive modeling, formal analysis of psychotherapeutic interventions, and verifiable embedding of safety constraints. To our knowledge, this is the first mathematically rigorous yet engineering-practical foundation for transdisciplinary cognitive science.

This study addresses computational psychology by developing a predictive, interactive, and deployable framework for psychological state modeling and analysis. To tackle instability in affective numerical prediction, we propose a stabilized Transformer-based regression model; for resource-constrained environments, we design a parameter-efficient fine-tuning strategy (LoRA) and a microservice-oriented deployment architecture. Our full-stack technical pathway integrates benchmark datasets, robust modeling techniques, and generative dialogue integration. Key contributions include: (1) the first reproducible, large-scale, democratized methodology for AI-driven psychological research; (2) empirical validation of prediction robustness across four major psychological datasets; (3) end-to-end coupling of predictive models with a personalized generative dialogue system (“Personality Brain”); and (4) a highly scalable, production-ready psychological analytics service platform.

Current human behavior modeling in educational and training contexts suffers from insufficient psychological fidelity and limited interpretability. To address this, we propose a multi-agent psychological simulation system featuring a novel “Inner Council” mechanism—comprising heterogeneous agents representing core psychological constructs (e.g., self-efficacy, mental models, social construction) that collaboratively generate transparent, theory-grounded decisions. The system integrates established learning theories—including social learning theory and cognitive apprenticeship—to support high-fidelity, traceable behavioral generation. Empirical evaluation demonstrates significant improvements in simulating higher-order learning mechanisms—such as metacognition and deliberate practice—in teacher training and psychological research. Crucially, our approach overcomes key limitations of black-box neural network models by ensuring psychological coherence and behavioral interpretability, thereby advancing the rigor and applicability of computational models in educational psychology.

This study addresses the dynamic modeling of mental health states by integrating the interplay between individual traits and situational factors. Drawing on interactionist and constructivist psychological theories, the authors leverage longitudinal social media data to extract situational linguistic features using the Situational Eight DIAMONDS framework, which are then combined with psychometrically informed large language model embeddings to build an interpretable predictive model. Results demonstrate that this theory-driven approach achieves competitive performance in predicting well-being while substantially enhancing model interpretability. Qualitative analyses further corroborate the alignment of key predictive features with established psychological theory, underscoring the validity and theoretical grounding of the proposed methodology.

Traditional user modeling approaches implicitly handle psychological states, limiting their ability to accurately interpret behavior in long-term, socially interactive settings. This work proposes the Mind Modeling (M3) framework, which for the first time systematically integrates Theory of Mind (ToM) into user modeling by explicitly representing mental states such as beliefs, intentions, emotions, and knowledge. M3 employs an integrated perception–mentalization–action architecture that enables dynamic inference and continuous updating of these states. The approach significantly enhances the interpretability and cross-session consistency of personalized systems. Feasibility is demonstrated through embodied interaction trajectories, establishing M3 as a novel paradigm for next-generation personalization.

Current research on Machine Theory of Mind lacks a rigorous formal definition and a unified evaluation framework. This work addresses this gap by integrating empirical principles from cognitive psychology, neuroscience, and artificial intelligence to propose the first formal definition of Machine Theory of Mind. It introduces a metamodel that synthesizes cognitive modeling, formal logic, and systems analysis, accompanied by a corresponding empirical benchmarking framework. By establishing a clear theoretical foundation, a unified modeling paradigm, and an actionable evaluation methodology, this study advances the field toward systematicity and empirical verifiability.

This study addresses a key challenge in developing socially trustworthy artificial intelligence: constructing human-like agents that exhibit both stable personality traits and adaptive behavior across diverse social contexts. The authors propose a novel architecture integrating the Big Five personality model with Bourdieu’s theory of cognitive–social co-construction, comprising an Individual Structure (IS) and a Multi-Scenario Contextual (MSC) framework. By leveraging structured prompts, the approach guides small language models to generate responses that are both personality-consistent and contextually appropriate. The method innovatively couples psychological personality modeling with sociological contextual structures through structured agent profiles, role–relationship–norm-based scenario modeling, and fixed prompt binding. Empirical results demonstrate significant improvements over non-fine-tuned large-model baselines in personality consistency, contextual adaptability, and stylistic alignment, while ablation studies confirm the essential contributions of both IS and MSC components.

This work proposes a psychologically grounded computational architecture centered on intrinsic needs, framing artificial general intelligence as an optimal decision-making problem aimed at fulfilling existential requirements within uncertain environments. By formalizing psychological mechanisms into a state space encompassing needs, perception, and action—and integrating experiential learning—the framework enables need-driven autonomous decision-making. Innovatively conceptualizing psychology as the agent’s “operating system,” the approach constructs a minimal viable model that jointly optimizes goal achievement, survival risk mitigation, and energy efficiency. Empirical validation demonstrates the feasibility and effectiveness of this need-oriented paradigm for intelligent decision-making.

This study addresses the lack of a unified framework for explaining the emergence of social relationships and social intelligence in long-term human–AI interaction. Modeling human–AI dialogue as a self-organizing socio-cognitive system, this work integrates affective adaptation, relational structuring, social memory, and personality consistency to propose novel theoretical constructs—including multi-timescale cognition, relational attractors, trust basins, developmental phase transitions, and socio-cognitive energy dynamics. Leveraging 14,700 dialogue turns and combining dynamical systems modeling with theory-driven empirical analysis, the research reveals hierarchical temporal persistence in social cognition, stable relational attractors, phase-transition-like developmental patterns, and a structured energy landscape. Notably, social intelligence exhibits a significant negative correlation with cognitive energy (r = –0.391, p < 0.001), and interactions demonstrate a consistent trend of energy decay over time.

This study addresses a critical limitation in current autonomous agents—their frequent failure to appropriately determine when, why, and whether to intervene—by proposing an explanatory model that integrates scene context, situational factors, and human behavioral considerations. For the first time, this work bridges humanities and engineering perspectives to develop a user-meaning-centered framework for behavioral explanation. The model explicitly distinguishes between observable facts and the contextual meaning ascribed by users, leveraging this distinction to formulate timely and proportionate intervention strategies. Furthermore, the research articulates five design principles for behavior judgment in intelligent agents, establishing a theoretical model and corresponding design guidelines that enhance contextual sensitivity and behavioral reasoning. This approach significantly improves the appropriateness of AI interventions and user acceptance.