Wilson John Sterking LAURET
- Chercheur indépendant
Présentation
Independent researcher specializing in crowd behavior and collective dynamics, with a focus on endogenous stability, governability, and irreversible transitions in complex systems.
Author of the Crowd-Based Dynamics (CBD) framework, including its foundational laws, formal mathematical structures, and diagnostic regimes. Research activity spans collective psychodynamics, entropic and mimetic failure modes, and reasoning architectures for Artificial Intelligence, including the formalization of RAG–RES–CBD interactions.
Work is characterized by a strong emphasis on theoretical closure, epistemic rigor, and non-normative governance models. Research Areas : Crowd-Based Dynamics (CBD); Collective behavior and mimetic processes; Endogenous governability and irreversible transitions; Entropic overload and saturation phenomena; Non-saturated mimetic rigidification; AI reasoning architectures and stability; RAG–RES–CBD formal modeling; Topological and geometric approaches to collective dynamics; Thresholds of irreversibility and systemic collapse Core Contributions : Reasoning in Artificial Intelligence and RAG–RES–CBD
Formal analysis of Reasoning-Augmented Generation (RAG) and Response Alignment Generation (RES) architectures as collective cognitive processes.
Integration of RAG and RES within the extended CBD framework.
Study of mimetic reinforcement, entropic overload, and loss of governability phenomena in large-scale AI systems.
Contribution to the formalization of an endogenously constrained AI (ECA) focused on diagnosis and stability. Crowd-Based Dynamics (CBD) Framework :
Formalization of CBD as a closed theoretical framework for collective dynamics.
Definition of foundational laws governing accumulation, saturation, and reversibility.
Development of a unified structure linking psychodynamics, temporal governability, and structural thresholds. Foundational Laws and Limits :
Law of Mimetic Saturation
Law of Conditional Reversibility
Informational Tipping Law
Temporal Governability Law
Sterking Limit and Universal Stability Limit These laws collectively describe the conditions under which collective systems remain governable or enter irreversible regimes. Key Theoretical Developments
Entropic Regime
• Formalization of the entropic bug as a loss of governability caused by exogenous overload.
• Correct identification of the small parameter ε(t)=K(t)/∣RES−RAG∣.
• Strict separation between local analytical approximation (DL) and global irreversible rupture.
• Preservation of the Sterking threshold (~2.14) as a global inequality, not an equality. Non-Saturated Mimetic Regime
• Formalization of the non-saturated mimetic bug as a silent, progressive loss of governability.
• Introduction and rigorous definition of the effective variety observable Ω(t).
• Canonical criterion :
Ω˙(t)<0 avec M˙(t)>0 et Dr≈0
• Demonstration that stability and performance may persist while plasticity collapses. AI Reasoning and RAG–RES–CBD Modeling
Formal analysis of Reasoning-Augmented Generation (RAG) and Response Alignment Generation (RES) as collective cognitive processes.
Integration of RAG–RES dynamics into the CBD framework.
Study of AI stability, mimetic reinforcement, and entropic overload in large-scale reasoning systems.
Contribution to the theory of endogenously constrained AI (ECA). Mathematical and Conceptual Tools :
Wasserstein geometry and optimal transport
Gradient flows in probability spaces
Topological quotient constructions (TCFQ)
Diagrammatic categorical representations
Differential and global regime separation
Diagnostic metrics vs. foundational laws Publications and Manuscripts (Zenodo-HAL-OpenAIRE-Academia)
https://orcid.org/0009-0004-7889-6641
Crowd-Based Dynamics (CBD) — Core Framework and Foundational Definitions
The Informational Tipping Law (Pure CBD)
Temporal Governability Law
Formal Framework of Collective Psychodynamics and Mimetic Thresholds
The Sterking Limit, Stability, and Governance
Beyond Generative AI: The Science of Endogenous Stability
Definitive Formalization of the Entropic Regime (Doc 1)
Formalization of the Non-Saturated Mimetic Bug (Doc 2) Research Positioning :
Independent, non-institutional research.
No normative optimization or exogenous control models.
Emphasis on diagnosis, structural understanding, and epistemic closure.
Framework designed to be theoretically closed and operationally open.
Compétences
Publications
Publications
A Foundational and Closed Framework for Collective Stability, Saturation, and Temporal Governability2026 |