Research
Independent research across six domains.
The Limits of My Tokens: The Token-Substrate Hypothesis and the Coinage Probe
We argue that for a large language model (LLM) the externally writable in-context token sequence IS the cognitive substrate for category-use — not a representation OF cognition that runs on some deeper substrate, but the substrate itself, the only handle that is movable from outside the weights.
Cite (BibTeX)
Mares, A. (2026). The Limits of My Tokens: The Token-Substrate Hypothesis and the Coinage Probe (Version 1.0.0) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.20157153
@misc{mares2026tsh,
author = {Mares, Alexandru},
title = {The Limits of My Tokens: The Token-Substrate Hypothesis and the Coinage Probe},
year = {2026},
publisher = {Zenodo},
version = {1.0.0},
doi = {10.5281/zenodo.20157153},
url = {https://doi.org/10.5281/zenodo.20157153}
}Research program
Elastic Automators
A diagnostic vocabulary for language-model-driven workflow systems.
Token-Substrate Hypothesis
Position paper paired with a pre-specified multi-model probe study; argues that for an LLM the in-context token sequence IS the cognitive substrate for category-use.
Textual Kinematics
Kinematic analysis of text as a dynamical system; AI text detection via temporal physics.
Yon Notation
A constructed notation for AI cognition; grammar and vocabulary decoupled.
Sightline Encoding Notation
A compiled reading format for non-linear cognitive consumption.
Sai (才) — Synthetic Autonomous Intelligence
A class of cognitive system designed to grow alongside a single human handler over a decadal timescale.