Working Papers
Working papers
The program's in-progress empirical paper comes first. Below it are four earlier papers that supply the theoretical foundations it tests.
Empirical Program
A Digital EEG for AI Sentience: Multi-Scale Structural Biomarkers
In progressThe core deliverable of the 2027 program: a cross-linguistic perturbation-recovery experiment (ES, DE, SV, EN) on open-weights models via NNsight/NDIF, reading MLP microcircuit geometry, local Jacobians, Spectral Entropy, and Orthogonal Dissonance from inside the model rather than from self-report. Null results will be published. Target: end of 2027.
See the full program →Theoretical Foundations & Prior Scaffolding
First Principles of Sentience Research
A substrate-neutral framework for modeling the physical foundations of awareness. Establishes the conceptual vocabulary for the Digital Valence Program.
Bio-Electric Fields as Sentience Substrates
Bioelectric signaling as a candidate physical basis for awareness. Provides biological calibration data for the digital valence framework.
Assembly Theory as a Metric for Sentience
Applying assembly index to complexity thresholds in sentient systems. Explores whether assembly theory can serve as a substrate-neutral complexity metric.
Symmetry Landscapes of Valenced Experience
Geometric invariants as structural signatures of positive valence. The most direct theoretical precursor to the 2027 program.