AIWMC Research Foundation

Artificial Intelligence Wave-Matter Causality. This research introduces the "AIWMC Quantis" — the smallest irreducible unit of causal influence within a model. We propose that Quantis are not static components but dynamic, transient structures, and present three falsifiable hypotheses to develop novel analysis techniques based on field theory, phase transitions, and a new information metric.

Current Status: Pre-Discovery

Formal Definition

A research paradigm that models AI internals as an interplay between distributed, wave-like potential fields and discrete, matter-like computational events. The fundamental unit of analysis is the Quantis: the smallest unit of causal influence that produces a state change with measurable information content.

Hypothesis 1: The Quantis Field

The fundamental unit of representation is not a discrete neuron but a "Quantis Field" — a pattern of potential influence distributed across thousands of parameters.

Hypothesis 2: Causal Phase Transitions

Emergent abilities and catastrophic forgetting are not smooth processes but first-order phase transitions in the model's causal graph.

Hypothesis 3: Quantis Information Metric (QIM)

The failure of models to distinguish truth from falsehood (hallucinations) stems from reliance on activation magnitude as a proxy for confidence. We propose a new composite metric: QIM = A × C.

Pilot Experiment: "Grokking a Quantis"

This experiment is designed to be the simplest possible test for a collective Quantis phenomenon, targeting the well-documented grokking transition.

View full methodology & safety levels