Procela: Epistemic Governance in Mechanistic Simulations Under Structural Uncertainty
Summary: arXiv:2604.00675v1 Announce Type: cross
Abstract
Mechanistic simulations typically assume fixed ontologies: variables, causal relationships, and resolution policies are static. This assumption fails when the true causal structure is contested or unidentifiable—as in antimicrobial resistance (AMR) spread, where contact, environmental, and selection ontologies compete. We introduce Procela, a Python framework where variables act as epistemic authorities that maintain complete hypothesis memory, mechanisms encode competing ontologies as causal units, and governance observes epistemic signals and mutates system topology at runtime. This is the first framework where simulations test their own assumptions.
Introduction
The understanding of complex systems has evolved significantly, particularly in fields like epidemiology and ecology. Traditional mechanistic simulations often rely on a fixed set of assumptions regarding causal relationships and variables. However, this rigidity can lead to inaccurate models, particularly in situations where the underlying structures are uncertain or debated. The emergence of antimicrobial resistance (AMR) serves as a prime example where competing ontologies must be considered to accurately capture system dynamics.
Introducing Procela
Procela represents a significant advancement in the field of mechanistic simulations. The framework introduces several innovative features:
- Epistemic Authorities: Variables within Procela act as epistemic authorities, maintaining a comprehensive memory of hypotheses related to system dynamics.
- Causal Units: Competing ontologies are encoded as causal units, allowing the simulation to reflect real-world complexities and uncertainties.
- Adaptive Governance: The governance mechanism observes epistemic signals and can mutate the system topology in real-time, responding dynamically to changes and uncertainties.
Application to Antimicrobial Resistance
To demonstrate Procela’s capabilities, the framework was instantiated for a hospital network grappling with three competing families of AMR. The governance mechanism was tasked with detecting coverage decay and policy fragility while running structural probes to assess the performance of various strategies.
Results
The results of the experiments were promising, showcasing significant improvements in the accuracy and adaptability of the simulations:
- Error Reduction: The framework achieved a 20.4% reduction in error rates compared to traditional baseline models.
- Cumulative Regret Improvement: There was a notable 69% improvement in cumulative regret over baseline simulations, highlighting the effectiveness of Procela’s adaptive governance.
Conclusion
Procela establishes a new paradigm in the realm of mechanistic simulations by effectively modeling not only the external world but also the internal processes of the modeling itself. This dual capability allows for greater adaptability under conditions of structural uncertainty, paving the way for more accurate and responsive simulations in complex systems like AMR.
Reproducibility and Auditability
All experiments conducted using Procela are fully reproducible, ensuring transparency and auditability in the simulation process. This commitment to reproducibility underscores the framework’s potential to become a foundational tool in the study of complex systems.