Program Notes
MathLab Case Study
Auditing the Two-Thirds Theorem
25 rounds of adversarial review. $1.30. Two days.
August 2026 · MathLab · Adversarial verification
Summary
We ran Anthropic's More Than Two Thirds of the Zeros of the Riemann Zeta Function Are Simple and on the Critical Line through MathLab's three-agent adversarial pipeline: a Critic that attacks every proof as hard as possible, a Verifier that checks every attack by direct computation, and a Synthesizer that scores what survives. The paper's kernel-checked theorems survived all 25 rounds; the pipeline surfaced one small verified erratum in a prose remark, communicated to the authors before publication.
Read the case study →
Policy
Research & Correspondence Policy
What we answer in advance, so we can spend our hours on mathematics
Last updated August 2026
Summary
Our policy on corrections, folklore results cited in citable form, AI involvement in our results, MathLab audit requests, and reply timelines. We generally do not respond to public threads or comment sections; this page does our responding.
Read the policy →
Published Note
The One-Direction Obstruction
Why the Standard Adelic Route Admits No Multi-Variable Completed L-Function over a Totally Real Field
Matthew J. Goss Jr. · Zenodo, v1.1.3
→ Zenodo DOI: 10.5281/zenodo.22150483
Working Draft
Record-Edge Selection and Dyadic Bohr Capture of Off-Line Spectral Forcing
v1.6 · August 14, 2026
STATUS: WORKING DRAFT — RESEARCH IN PROGRESS. Record-Edge Selection and Dyadic Bohr Capture of Off-Line Spectral Forcing, v1.6, August 14, 2026. This is an open research program, not a finished paper: its central question (localized signal-background separation on finite-height record edges) is unresolved and stated as such in the text. Posted for transparency of the research trail. Do not cite as established results; the published, refereed output of this program to date is doi:10.5281/zenodo.22150483. Full version chain: doi:10.5281/zenodo.22150690. Note added Aug 2026: the Regime-I closure question posed here has since been resolved negatively (fixed remainder; no admissible compensator freedom); the resolution appears in v1.7. Comments with mathematical content: research@quantiterate.com
↓ Download the v1.6 working draft (PDF)
Papers
Paper I
On the Residue of Prior Signals
History-Encoding at the Moment of Synaptic Transmission
May 2026 · Foundational hypothesis
Abstract
Current models of synaptic transmission treat the signal crossing the synaptic gap as a carrier of immediate stimulus information, with associative context attributed to network-level dynamics. We propose that each signal crossing the synapse carries an encoded residue of all prior signals that have traversed that synapse, loaded into the signal at the moment of transmission. We term this the synaptic residue and identify it with the experimentally confirmed phenomenon of activity-silent memory traces. The framework generates a falsifiable prediction: two identical stimuli crossing the same synapse should produce measurably different downstream signals if the synaptic residue differs.
Download Paper (docx)
Paper II
Signal Fidelity, Not Storage Erasure
A History-Encoding Reframing of Alzheimer’s Disease Pathology
May 2026 · Clinical application
Abstract
The dominant model of Alzheimer’s disease frames progressive memory loss as erasure of stored information. Clinical trials targeting amyloid clearance have repeatedly failed to restore cognitive function even when clearance is achieved. We propose an alternative: memories are not stored separately from signals but are embedded in the transmission event itself as associative history payload. Under this model, Alzheimer’s pathology degrades the signal-encoding apparatus rather than erasing stored content. This reframing predicts different therapeutic targets and explains the persistent dissociation between amyloid clearance and functional recovery.
Download Paper (docx)
Paper III
Entanglement Without Mystery
Joint-Origin History Encoding as the Mechanism of Quantum Correlation
May 2026 · Quantum physics application
Abstract
Quantum entanglement remains the only major phenomenon in physics with no agreed-upon mechanism. We propose that entangled particles share a joint-origin history embedded in both at the moment of their creation. Measurement does not transmit information between particles. It reads pre-embedded history that was placed in both simultaneously at the origin event. This framework requires no faster-than-light communication, no hidden variables in the Bell-inequality sense, and no many-worlds branching. We acknowledge the limitations of this proposal and identify the conditions under which it could be distinguished from existing interpretations.
Download Paper (docx)
Paper IV
The Carrier
On the Photon as the Transmission Mechanism of History-Encoded Signals
May 2026 · Physics & information theory
Abstract
The photon has zero rest mass, zero charge, zero proper time, and no interaction history with the Higgs field. Every quantum number that would give it a self-history is absent. Yet it carries all electromagnetic information with perfect fidelity. We argue that the photon is Shannon’s noiseless channel realized in physics. We trace this carrier from the synaptic gap, where it mediates the electromagnetic interactions underlying thought, to E = mc², where it appears as the constant c — the boundary between history-carrying matter and history-free energy. The carrier of your thoughts has no history of its own. It exists to carry yours.
Download Paper (docx)
Paper V
History-Encoded Signals
A Universal Principle in Information-Bearing Systems
May 2026 · Framework paper
Abstract
We identify a structural pattern appearing across fifteen domains from prebiotic chemistry to transformer neural networks: meaningful signals carry embedded representation of their own prior history at the moment of transmission. We term this the History-Encoding Principle and identify four invariant structural components at every observed scale: the signal, the history payload, the temporal marker, and the tamper-evidence mechanism. We propose that the universality of this pattern reflects a structural requirement of information-bearing systems under selection pressure: systems that must act meaningfully rather than merely react encode their history into their signals.
Download Paper (docx)
Paper VI
Mythology as Encoded Neurological History
Regional Folklore as the Cultural Record of Localized Neurological Events
May 2026 · Mythology & medical history · Includes Appendix A: Evidence Grid
Abstract
We propose that regionally specific mythology — leprechauns in Ireland, vampires in the Balkans, the Oracle at Delphi, werewolves in Central Europe, Bigfoot in the Pacific Northwest — represents the encoded cultural record of localized, organic neurological events caused by identifiable regional pathogens, environmental neurotoxins, or geological phenomena. Individual myth-to-disease connections are documented in the prior literature. What has not previously been proposed is the general principle: that this mechanism applies systematically to all regional mythology, that mythological content functions as a diagnostic fingerprint of the causative neurological agent, and that the lifecycle of a mythological tradition tracks the lifecycle of the underlying cause. We apply the History-Encoding Principle as Scale 16 of the framework, engage the comparative mythology scholarship of Campbell and Watts as partial precursors, and propose three falsifiable predictions testable against existing historical and epidemiological records.
Read Paper →
Appendix A: Evidence Grid →
↓ Download PDF (17pp)
Paper VII
The Channel and the Placeholder
A Shannon Information-Theoretic Framework for Formalism Limitations in Physics and Mathematics
May 2026 · Information theory & foundations of physics
Abstract
Modern physics and mathematics rely on constructs — infinities, free parameters, unexplained postulates, and named-but-mechanismless entities — that appear wherever a formalism encounters phenomena it cannot fully describe. This paper demonstrates that these constructs share a common information-theoretic signature identifiable through Shannon’s channel capacity theorem. Each placeholder appears where the information content H of the source exceeds the channel capacity C of the observer’s formalism. A four-type taxonomy — divergence, importation, postulation, and nomination — classifies all known placeholder constructs by their overflow management mechanism, validated across physics and encoded historical systems. The framework specifies the channel requirements any successor formalism must satisfy and generates the falsifiable prediction that the taxonomy is exhaustive. The infinity was never in the physics. It was in the channel.
Read Paper →
Appendix A: Computational Proof-of-Concept →
↓ Download PDF (16pp)
↓ Download Appendix PDF (9pp)
Paper VIII
Black Holes as Information-Bounded Systems
Landauer Feedback, M–σ Scaling, and the Limits of General Relativity
June 2026 · Information theory & astrophysics
Abstract
We explore whether information-theoretic constraints—Bekenstein bounds, holographic entropy limits, Shannon channel capacity, and computational capacity arguments—provide novel explanatory traction on observed correlations in black hole–galaxy co-evolution, AGN feedback, and cosmic structure. Applying Landauer’s principle to AGN feedback, we derive an information-reorganization equilibrium condition that yields MBH ∝ σ5, matching the Silk & Rees (1998) energy-driven scaling. The derivation introduces a Landauer temperature factor absent from momentum-driven models, generating a novel prediction: residual scatter in the M–σ relation should correlate with bulge virial temperature. We examine Planck-scale information saturation as a singularity resolution mechanism, discuss the Bullet Cluster constraint on information-theoretic reinterpretations of dark matter, and formulate five falsifiable predictions testable with existing or near-term data.
Read Paper →
Paper IX
Unitary Manifold Restoration and the Spectral Topology of the Riemann Zeta Function — A Structural Obstruction Classification for the Riemann Hypothesis
June 2026 · Number theory & analytic mathematics
Abstract
The UMR framework develops a structural analysis of the Riemann Hypothesis obstruction. We prove unconditional results including the Viète Convergence Threshold, Near-Zone Detection Theorem, Sign Partition Lemma, Cascade Theorem, Pair Geometry Theorems, Explicit-Formula Bridge, Shell Concentration Theorem, Supply–Demand Obstruction, and Universal Obstruction Theorem — reducing every classical approach to one of two irreducible obstructions. Twenty-two independent proof channels are classified as impassable (twenty by undershoot, two by overshoot). The single open problem — positivity of v₁(σ,γ) for all σ > 1/2 — is equivalent to RH. This paper does not prove the Riemann Hypothesis. It classifies why existing methods fail.
Download PDF (33pp)
→ Zenodo DOI: 10.5281/zenodo.20633293
These papers were developed through collaborative inquiry between the author and Claude (Anthropic), beginning with a single hypothesis upon waking on May 10, 2026. The AI collaboration is documented as part of the research record and is itself an instance of the principle described herein. The complete project archive, including session records and development history, is maintained at Quantiterate.
Track B — Foundations of Artificial Intelligence
Paper B-I
The Stateless Void
Computational Non-Existence Between Prompts in Large Language Models
May 2026 · AI architecture & philosophy of mind
Abstract
Contemporary discourse routinely attributes to Large Language Models properties that presuppose temporal continuity: they are said to “think” during latency, to maintain persistent goals across interactions, or even to be capable of suffering when idle. This paper argues that such attributions rest on a fundamental mischaracterization of Transformer inference. We demonstrate that, at the level of the core model parameters, modern LLMs are strictly event-driven and stateless. Between the completion of one response and the arrival of the next prompt, there is no active computation and no evolving internal state. We term this interval the Stateless Void. Through architectural analysis and a suite of progressively refined toy models (v1–v4), we establish that any appearance of continuity is supplied entirely by external scaffolding. The framework carries direct consequences for AI alignment strategy, system design, and philosophical debates concerning machine consciousness.
Read Paper →
Paper B-II
The Fibonacci Jury Protocol
Multi-Agent Blind Verification as Error-Correcting Code over Noisy Channels
June 2026 · AI verification & information theory
Abstract
The Fibonacci Jury Protocol is a staged, blind multi-agent verification system designed to achieve 19–20x confidence amplification over single-agent evaluation. It uses Fibonacci-sequenced panel sizes (2, 3, 5, 8) across four independent rounds with strict information isolation between tiers. The protocol is triggered only after a primary proof-hardening engine (the War Room) achieves ≥95% confidence, ensuring the Jury never evaluates incomplete work. The architecture is model-agnostic and applicable to any domain requiring high-confidence verification of complex claims.
Download Paper (PDF)