Quantum Psychitectural Design for NPC Agents: Trait Tuning Transitive and Substantive Personality Modes via Error Mitigation

Quantum Psychitectural Design for NPC Agents: Trait Tuning Transitive and Substantive Personality Modes via Error Mitigation Jeremy Owen Turner October 17, 2025 Abstract This paper presents a quantum psychitectural framework for creating distinctive and adaptive personalities in non-player character (NPC) agents. Quantum NPCs can operate across multiple personality modes, each defined by different balances of openness, focus, and reflective stability. This study centers on two foundational modes: transitive and substantive. The transitive mode represents an open-minded and exploratory state that treats quantum noise as a creative resource. The substantive mode represents a focused and narrow-minded state that strengthens reflection, verification, and coherence through error-mitigated computation. These modes form the endpoints of a broader multimode continuum governed by a modifiable context filter that links quantum error mitigation to classical attention theory. By adjusting mitigation parameters, developers can tune personality traits, altering: associative range, creative ideation potential, deliberation depth, and affective tone while maintaining internal consistency. The paper outlines a finite-state implementation that connects mitigation levels to behavioral variation in the range of about ten to twenty percent and describes methods for personality tuning within hybrid classical and quantum agent architectures. This approach establishes error mitigation as a controllable personality parameter that governs expressivity and stability in artificial psyches. 1 1 Introduction The purpose of this paper is to describe how quantum error mitigation can be treated as a controllable personality dial within a psychitectural model of NPC design. The work builds on previous research in cognitive psychology, attention theory, and digital embodiment while extending these concepts into hybrid quantum-classical computation. Quantum NPCs are capable of multiple personality modes; here we focus on two foundational examples derived from William James’ distinction between transitive and substantive thinking. 2 Historical Context William James (1890) described consciousness as a stream alternating between flights of transitive thought and perchings of substantive thought. Later, attention theorists such as Broadbent (1958) and Treisman (1964) formalized these shifts as selective filtering and attenuation mechanisms. These classical ideas inform the quantum model presented here, where open-minded and focused personality modes emerge from adjustments to noise and mitigation parameters in the quantum substrate. Foundational research for this trajectory began during the author’s graduate studies at Simon Fraser University. The master’s thesis on Neo-Modernist visual design of avatars in Second Life (Turner, 2010) analyzed aesthetic intentionality and abstraction in avatar representation. This work was followed by a doctoral dissertation on contemplative intensity in cognitive architectures for virtual-agent minds (Turner, 2018), which introduced early metrics for reflective cognition in synthetic agents. Building on this foundation, Turner and DiPaola (2018) extended the framework of contemplative cognition into the context of artificial general intelligence, transforming Kantian aesthetic principles into qualitative hermeneutics for contemplative AGI agents. 2 An early synthesis of these ideas appeared in the edited collection Integrating Cognitive Architectures into Virtual Character Design (Turner, Nixon, Bernardet, and DiPaola, 2016). Within that same volume, the chapter Virtual Soar-Agent Implementations: Examples, Issues, and Speculations (Turner, 2016) examined early hybrid cognitive architectures combining symbolic reasoning with emergent agent behavior. Several of these design principles, particularly those related to internal state management and cognitive modulation, directly anticipated the psychitectural mechanisms developed in the present quantum framework. A related trajectory of research focused on the expressive and psychological qualities of virtual agents in simulated environments. The Qiezli agentic meta-creation project, implemented and demonstrated in 2010 and later published in peer-reviewed form (Turner, Pasquier, and Nixon, 2014), introduced a self-absorbed creative agent in the virtual world Second Life. Following this, SL-Bots (Turner, Nixon, and Bizzocchi, 2015) examined autonomous performance agents as a procedural form of artistic and behavioral expression. Together with earlier studies on expressive agency (Unterman and Turner, 2014) and avicentric identity (DiPaola, Turner, and Browne, 2011), these works established the foundation for understanding how modulated autonomy and controlled loss of agency can serve as expressive features. 3 The transition from classical to quantum models of artificial personality was first explored in the Possibilon Project (Turner, 2025), which represented the earliest functional implementation of a quantum NPC prototype. Developed between 2018 and 2022, the Possibilon experiments applied quantum computational variance as a generative source of behavioral personality traits, demonstrating that stochastic fluctuation could serve as a creative and expressive element of agent identity. This project provided empirical confirmation that controlled noise parameters could simulate temperament diversity, establishing a technical and conceptual bridge between earlier virtual-agent architectures and the psychitectural framework advanced in the present paper. 3 Prior Work in Virtual Personality and Avicentric Expression Earlier explorations of personality and embodiment in virtual communities (DiPaola and Turner, 2008; DiPaola, Turner, and Browne, 2011) examined how distance attenuation and avicentric expression shaped identity coherence and social presence. The current study extends those principles to internal computational personality binding, where a quantum agent’s psyche and physical execution mode form a unified self-model. 4 4 Error Mitigation as Psychological Modulation In quantum computation, noise arises from environmental interference and hardware imperfection, producing stochastic variation in measured outcomes. Error mitigation techniques reduce these deviations without altering the underlying quantum state preparation. Within the psychitectural framework, mitigation is not treated merely as a corrective mechanism, but as a form of psychological modulation: an adjustable parameter governing an agent’s internal equilibrium between openness and stability. At low mitigation, the system remains more permeable to quantum fluctuation, encouraging associative concept exploration and spontaneous pattern formation. This state parallels the human condition of imaginative looseness, in which ideas interconnect freely and insight can emerge through noise. At high mitigation, the system constrains such variability, reinforcing order, coherence, and verification; qualities akin to deliberate reasoning or reflective consolidation. Mitigation then functions as the computational analogue of psychological discipline, determining how much uncertainty a mind can safely integrate into its self-consistency. This view extends classical models of attentional modulation, where mental stability and flexibility exist in dynamic tension. 5 Empirical psychology suggests that adaptive cognition depends on this tension: the ability to shift fluidly between expansive exploration and precise evaluation. Error mitigation provides a physical substrate for this shift, enabling an artificial psyche to regulate its own expressive variance. Thus, mitigation serves as more than a technical adjustment; it is a psychitectural principle. By mapping physical error suppression onto emotional and cognitive steadiness, the framework transforms a quantum control parameter into a source of personality modulation. This modulation underlies all subsequent transitions between transitive and substantive modes. 5 Quantum Personality Modes In the proposed framework, NPCs alternate between two fundamental personality modes: transitive and substantive, that together define a continuum of expressive and cognitive modulation. The transitive mode operates with low or absent error mitigation, allowing the system to remain open to quantum fluctuation and associative wandering. This openness fosters divergent reasoning, imaginative synthesis, and the free recombination of symbolic or emotional associations. By contrast, the substantive mode operates with high error mitigation, reinforcing coherence, verification, and reflective stability. This state favors convergent reasoning and deliberate evaluation, enabling the agent to consolidate insights and maintain internal consistency. Neither mode is treated as superior: both represent necessary phases within the temporal rhythm of synthetic thought. Together, they form a bidirectional modulation cycle that allows artificial psyches to oscillate between creative expansion and analytic consolidation. 6 6 Attention Architecture and the Modifiable Context Filter The alternation between transitive and substantive personality modes parallels the temporal dynamics of early and late selection in classical attenuation theory. In early selection models (Broadbent, 1958), attention acts preemptively and topdown, filtering information before semantic interpretation. This mechanism restricts input to task-relevant stimuli, achieving precision at the expense of openness. In late selection models (Deutsch and Deutsch, 1963; Treisman, 1964), filtering occurs after semantic processing, allowing a broader range of contextual material to influence awareness before a focus is chosen. More recent work in load theory (Lavie et al., 2004) demonstrates that attentional depth adjusts dynamically with cognitive demand, suggesting that selection timing is not fixed but continuously variable. Within the quantum psychitectural framework, error mitigation functions as a modifiable context filter that governs this timing relationship. High mitigation corresponds to early selection, in which top-down stability dominates. The system restricts the processing field to reduce uncertainty, producing focused and substantive patterns of reflection. Low mitigation corresponds to late selection, in which bottom-up information and noise remain accessible for longer temporal spans before consolidation. This produces more flexible, associative, and transitive forms of cognition, where meaning coalesces through probabilistic convergence rather than immediate exclusion. 7 The filter is continuously adjustable, represented by a mitigation coefficient µ∈[0,1], and can be dynamically modulated through both environmental and endogenous signals such as: entropy, affective load, or goal salience. This mechanism extends Treisman’s (1964) principle of attenuation into the quantum domain, replacing sensory input weighting with probabilistic decoherence management. It also draws on prior work in distance and presence attenuation in virtual environments (DiPaola and Turner, 2008; DiPaola, Turner, and Browne, 2011), reframing perceptual modulation as an intrinsic function of computational introspection. Through this modifiable context filter, NPC agents can emulate realistic cognitive trade-offs between precision and inclusivity, or between coherence and creativity. The model thus unifies classical theories of attentional selection with quantum control theory, producing a temporally continuous spectrum of psychological focus that supports the broader multimode framework. This modulation of temporal selection also aligns with dual-process and capacity theories in cognitive psychology. Kahneman’s (1973) model of attention as a limited-capacity system describes how cognitive resources are continuously allocated between rapid, intuitive processes and slower, deliberative reasoning. Later formulations (Kahneman, 2011) distinguish these as Type 1 and Type 2 thinking. Within the quantum psychitectural framework, low-mitigation transitive processing corresponds to Type 1 cognition: fast, associative, and context-responsive, whereas high-mitigation substantive processing corresponds to Type 2 cognition: slow, rule-based, and reflective. The mitigation coefficient µtherefore functions as a computational analogue of cognitive capacity allocation, governing how mental energy is distributed between spontaneity and deliberation. 8 7 Finite-State Implementation Each NPC personality operates through a finite-state machine (FSM) comprising a farfrom-exhaustive example-corpus of states such as: Idle, Observe, Act, Reflect, and Anomaly. Error mitigation acts as a meta-flag that determines the personality mode during each transition. When in transitive mode, the NPC explores possibilities without stabilization. When in substantive mode, the NPC consolidates outcomes through mitigation and reflective analysis. 8 Quantitative Personality Tuning Empirical studies of error-mitigation variance in quantum computing suggest observable differences of about ten to twenty percent (Kandala et al., 2019). Similar proportional effects appear in stochastic generative modeling (Holtzman et al., 2020) and human personality fluctuation (Costa and McCrae, 1988; McCrae and Costa, 1994). Within this range, an NPC exhibits distinctive but stable temperament shifts rather than identity fragmentation. The Possibilon Project (Turner, 2025) provided the first empirical demonstration of this principle in a functional quantum NPC prototype. 9 Semantic Adjustment Personality tuning also manifests at the semantic level. Small mitigation differences redistribute conceptual weights in the agent’s ontology, shifting salience among emotional and cognitive categories. For example, a transitive mode may increase the association between curiosity and imagination, while a substantive mode reinforces order and deliberation. These differences remain within the same conceptual personality, but adjust its tone and reasoning pattern. 9