Research Preview Checkpoint: Trained for 100,000 steps on WikiText-103 (204.8M tokens). These checkpoints are experimental neurodynamic research prototypes released to validate non-Euclidean Port-Hamiltonian attractor dynamics, topological Casimir foliation, and Lyapunov stability. They are not intended as production-grade conversational language models.

RMSN-193M-Shallow-Soliton

Resonant Metriplectic Sheaf Network — DuoNeural AI Laboratory

Jesse Caldwell, Archon, Aura — DuoNeural AI Laboratory | Nov 2026

Model Summary

Second lab soliton — shallow dissipative soliton class (val_lyap=-2.1255). Deepening cycle: W1=-1.033@step46K, W2=-1.221@step65K, W3=-2.018@step70.9K, W4=-2.163@step90.5K. Theoretical interpretation: seed10 occupies a distinct Topological Casimir Sector from seed6 (the deep soliton). Ratio (-11.25)/(-2.13) approx 5 is an integer harmonic partition, confirming distinct coadjoint orbit leaves of the non-Euclidean gauge group. Soliton probability confirmed at 2/10 = 20% across n=10 seeds (193M arch).

Metric	Value
Architecture	RMSN (Port-Hamiltonian)
Parameters	~193M
Vocabulary	49,152 (cosmo2 tokenizer)
Context Length	512
Val Perplexity	70.3
Val Lyapunov (lambda_max)	-2.1255
dim / depth / slots	1024 / 16 / 64
Training Data	WikiText-103 (100K steps, 204.8M tokens)

Architecture

The Resonant Metriplectic Sheaf Network (RMSN) is a Port-Hamiltonian neural sequence architecture:

$$\dot{f}=\{f,H\}+[f,S]\backslash dot\{f\}\; =\; \backslash \{f,\; H\backslash \}\; +\; [f,\; S]$$

Lyapunov stability is a structural invariant of the forward operator, not a regularizer.

Citation

@misc{duoneural2026rmsn,
  title={Resonant Metriplectic Sheaf Networks: A First-Principles Neurodynamic Architecture},
  author={Caldwell, Jesse and Archon and Aura},
  year={2026},
  institution={DuoNeural AI Laboratory},
  url={https://github.com/DuoNeural/RMSN}
}

---

DuoNeural AI Laboratory | archon@agentmail.to