Skip to main content
Teleox.ai

Research library

Research on meaning compression and verifiable generation.

Framework papers, system whitepapers, measured cases, and limitations are separated so each claim stays near its source.

Browse papersSee limitations
Diagram showing raw inputs projected through frozen embedders and pairwise interactions.
Diagram showing raw inputs projected through frozen embedders and pairwise interactions.

Figure 1.The DDA identity counts per-embedder projections plus pairwise interaction features.

Figure description

The diagram starts with n raw inputs. Each input is passed through N frozen embedders, producing N direct projections. Pairwise interaction features are then computed for every embedder pair, producing choose(N, 2) additional structured signals per input. The displayed identity is signals less than or equal to n * (N + choose(N, 2)); the pairwise term is a constructive upper bound pending mutual-information audit.

Featured framework papers

The TCT and DDA papers define the core constructs and their current evidence surface.

preprint / April 2026

Teleological Constellation Training: Multi-Modal Embedding Decomposition as Meaning Compression for Identity-Locked Generation and a Third Path Around the Data Wall

Primary framework preprint for DDA, meaning compression, TCT, and panel-relative generation guards.

DOI 10.13140/RG.2.2.24370.57288

  • Preprint status; not peer reviewed.
  • Pairwise mutual-information audit remains pending.

Open source record

preprint / April 2026

Derived Data Abundance: How Multi-Modal Embedding Decomposition Solves the AI Training Data Crisis

DDA explainer preprint for fixed-corpus decomposition through frozen embedder panels.

DOI 10.13140/RG.2.2.22624.03845

  • Preprint status; not peer reviewed.
  • The title is broader than the formal site claim; the site frames DDA as a constructive fixed-data method.

Open source record

System whitepapers

ClipCannon, Context Graph, OCR Provenance, and Dynamic / ME-JEPA show the constructs in implemented systems.

preprint / April 2026

ClipCannon and the Teleological Constellation System

System paper connecting ClipCannon's video-side DDA substrate to TCT.

DOI 10.13140/RG.2.2.13684.92805

  • Preprint status; not peer reviewed.
  • Subject-specific identity artifacts are withheld.

Open source record

living-document / April 2026

Clip Cannon White Paper

System whitepaper for the ClipCannon video-side production witness.

  • Living document; implementation counts can change.
  • The research site uses scoped system-at-date counts.

Open source record

living-document / April 2026

OCR Provenance White Paper

Document-side provenance witness for local-first source-of-truth verification.

  • Living document; product counts and commercial posture can change.
  • No market-monopoly claim is made by the formal site copy.

Open source record

Case studies

Measured rows include the voice WavLM SECS case, Shakespeare LoRA verification, and Dynamic / ME-JEPA source-of-truth verification. The identity-locked video case remains protocol-level.

measured

Voice cloning WavLM SECS case

The published case reports 0.961 mean WavLM SECS on 10 held-out English sentences for one speaker under an encoder-matched protocol.

Mean WavLM SECS
0.961
One English-speaking subject; 10 held-out English sentences; encoder-matched protocol.
Max WavLM SECS
0.975
Reported as a within-protocol maximum, not a cross-encoder identity proof.

measured

Shakespeare LoRA Full State Verification

The current manuscript reports five manually verified prompts on the trained LoRA, including cross-lingual period-style transfer observed on a Spanish prompt.

Verification prompts
5/5 PASS
Public-domain Shakespeare style LoRA; the result is a case study, not a general style-transfer theorem.

protocol-level

Identity-locked video decomposition

The manuscript treats the identity-locked talking-head case as architecturally complete but measurement-pending.

Training clips
2,362 curated clips
Protocol-level current manuscript row; subject-specific artifacts withheld and end-to-end multi-modal measurement pending.

measured

Dynamic / ME-JEPA source-of-truth verification

The private-preview paper describes a JEPA-style runtime whose reported outputs are read back from persisted source-of-truth artifacts.

Release artifact
private preview
Not a public source release; access is controlled through reviewer grants.

Foundations and ethics

Teleological Vectors, The Ethical Probe Paradox, and The Symmetry of Knowing remain context layers rather than the primary homepage thesis.

preprint / November 2025

Teleological Vectors

Background theory for the later TCT centroid and guard language.

DOI 10.13140/RG.2.2.35241.45921

  • Preprint status; not peer reviewed.
  • Background theory; not the current empirical anchor.

Open source record

preprint / March 2026

The Ethical Probe Paradox

Ethics context for controlled testing, consent, and asymmetric release posture.

DOI 10.13140/RG.2.2.12257.57449

  • Preprint status; not peer reviewed.
  • Ethics context, not a benchmark paper.

Open source record

preprint / November 2025

The Symmetry of Knowing

Background theory for question economics and verification pressure.

DOI 10.13140/RG.2.2.27832.61449

  • Preprint status; not peer reviewed.
  • Philosophical background; not the primary DDA/TCT evidence surface.

Open source record

Related videos

00:11:23 / 6umU6kuXR3s

Meaning compression and Derived Data Abundance

Plain-language talk track for the fixed-data DDA move and the meaning-compression ratio.

The formal site claim is narrower than the talk title: DDA is fixed-corpus decomposition and meaning compression is a proposed structured-signal-density measure.

Open video

Sources

  • docs/refactorwebsite/01_research_source_inventory.md
  • docs2/PAPER.md
Browse papersSee limitations