Adaptive Domain Models: Bayesian Evolution, Warm Rotation, and Principled Training for Geometric and Neuromorphic AI

arXiv:2603.18104 · cs.AI, cs.DC, cs.LG, cs.NE · Submitted 2026-03-18 · Read on arXiv

cs.AI, cs.DC, cs.LG, cs.NE

Submitted: 2026-03-18

Updated: 2026-10-15

Comments: 26 pages, 0 figures, 2 tables. Major revision: expanded precision and efficiency analysis. Added Urschel's bounds and worked Bayesian examples

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Active Inference — 1× in this paper · explained in 1 episode
• Distillation — 1× in this paper · explained in 2 episodes
• Hypergraph — 1× in this paper · explained in 2 episodes

Sources

• Dimensional Type Systems and Deterministic Memory Management: Design-Time Semantic Preservation in Native Compilation
• The Program Hypergraph: Multi-Way Relational Structure for Geometric Algebra, Spatial Compute, and Physics-Aware Compilation
• Gradients without Backpropagation
• Beyond Backpropagation: Optimization with Multi-Tangent Forward Gradients
• Closing the Gap Between Float and Posit Hardware Efficiency
• Decidable By Construction: Design-Time Verification for Truly Fearless Systems
• Fixed-Point Scaffolding in the Clef Programming Language
• Clifford Group Equivariant Neural Networks
• Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
• HIF: The hypergraph interchange format for higher-order networks

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems