Artificial Intelligence and Artificial General Intelligence dynamic analysis of Travis Raymond-Charlie Stones models

Abstract 1:

Mackey-Glass Time Series Test on 16 Advanced Cognitive Models

This experiment was conducted to evaluate the performance of 16 advanced artificial intelligence and cognitive system models against the Mackey-Glass time series test an established benchmark for measuring a model's ability to predict and adapt to nonlinear, chaotic dynamics. The test simulates a delayed feedback system known for its sensitivity to initial conditions, making it an ideal metric for analyzing recursive adaptability, stability under chaos, signal alignment, and variance suppression.

The 16 models tested span a range of system architectures, including classical decay models, quantum-recursive hybrids, and higher-order artificial general intelligence (AGI) frameworks. Each model was run under identical conditions and evaluated on their ability to remain stable, track chaotic behavior, suppress variance, and exhibit forward-learning or feedback optimization.

Model Order in Results Table:

1. PSRS True  A baseline model with ideal harmonic decay and cosine behavior.

2. PSRS Overestimated  A model with higher amplitude and slower decay.

3. PSRS Underestimated  A more conservative variant with faster decay.

4. CRIS  Recursive feedback simulation with quantum tunneling logic.

5. CRISS  Enhanced CRIS with subconscious logic and bifurcation control.

6. CASI   A conscious-layer model for feedback stabilization and signal smoothing.

7. CRISS-CASI  An integrated subconscious-conscious hybrid system.

8. AGI-1  An early-stage recursive model with limited correction ability.

9. AGI-2  A feedback-tuned recursive learner with moderate stability.

10. AGI-3  A mid-stage recursive learner with improved signal matching.

11. AGI-4  Incorporates bifurcation anticipation and adaptive thresholding.

12. AGI-5  Adds probabilistic correction and spectrum tracking.

13. AGI-6  Enhanced short-term adaptability with balanced correction.

14. AGI-7  Adds noise suppression and phase-aware learning.

15. AGI-8  Integrates hierarchical reasoning and layered fallback logic.

16. AGI-9  Final model with recursive spectrum convergence and AGI-aligned decision structuring.

Each model's response was logged over 100 iterations, and performance was plotted and tabulated. Metrics evaluated include signal accuracy, alignment with chaotic attractors, phase synchronization, noise tolerance, and convergence stability. The goal was to identify which architectures exhibit the most intelligent, adaptive, and human-like pattern reasoning under dynamic and uncertain conditions.

The data resulting from these runs serves as a foundation for developing next-generation recursive intelligence systems capable of real-time adaptability and long-term coherence in chaotic environments.

Abstract 2:

Cognitive Recursive Integrated Simulation (CRIS )Test Evaluation Across 16 AI Models

This study evaluates sixteen advanced AI models using the CRIS (Cognitive Recursive Integrated Simulation) Test, a framework designed to assess recursive reasoning, adaptive feedback, and stability under complex signal environments. The models range from classical probabilistic simulations (PSRS) to hybrid quantum-classical systems (CRIS, CRISS, CASI, CRISS-CASI), and nine iterations of AGI (Artificial General Intelligence) prototypes.

The CRIS test simulates nonlinear cognitive feedback loops, incorporating spectrum convergence, bifurcation control, and quantum-inspired learning. Each model was evaluated based on its ability to maintain signal integrity, react to recursive deviation, and exhibit stabilization without overcorrection or oscillatory drift.

Key findings include:

CRISS-CASI demonstrated superior convergence and adaptability, ranking highest in dynamic regulation.

AGI-9 closely followed, with advanced recursive alignment and signal tracking accuracy.

CASI performed well in smoothing erratic deviations but lagged in rapid phase response.

Early AGI models (1-3) exhibited significant overreaction and underfitting under recursive stress.

PSRS variants provided a controlled baseline, clearly highlighting divergence tendencies in less adaptive systems.

This evaluation underscores the importance of recursive and spectral reasoning layers in managing complex, adaptive AI behaviors. The CRIS test confirms the trajectory toward integrated, conscious-subconscious hybrid architectures (e.g., CRISS-CASI) as the most promising direction for resilient, intelligent systems.

Let me know if you’d like it saved as a PDF or added to your final report.