Begin here

The sober map of TS.

TS is an engineering framework for modelling information transfer through constraint graphs, tension, propagation, relaxation, contradiction handling, and revision.

Public stack

TS-Core = graph/tension runtime.

TensionLM = sigmoid tension attention language model experiments.

CIG = persistent provenance-aware claim/evidence graph.

Proof Ranker = proof scoring, repair, and verifier loop experiments.

BoggersTheAI / Mind / Pulse = experimental runtime/agent branches, secondary to the public core stack.

Docs
Proof Bank
Website
CIG
TensionLM
Proof Ranker
TS-Core graph/tension substrate
Propagate
Relax
Break
Evolve

What TS is

Engineering framework

Experimental AI architecture lens

Graph/tension/provenance/reasoning substrate

A way to make internal model behaviour more inspectable

What TS is not

Not a finished AGI claim

Not a theory-of-everything proof

Not financial advice

Not claiming benchmarks are beaten until receipts exist

Not replacing normal scientific evidence

Where to go next

Pick the route that matches the kind of evidence you want first.

For researchers

Start with research framing and proof-bank receipts.

First route Second route

For engineers

Inspect project structure, source repos, and replay paths.

First route Second route

For funders/collaborators

Read the grounded bio, roadmap, and contact route.

First route Second route

For casual readers

Read the plain-language primer and field notes.

First route Second route

Core project nodes

These are the public routes most claims should pass through.

TS-Core

Minimal graph/tension runtime for inspectable constraint propagation and relaxation.

Open node

TensionLM

Sigmoid tension attention experiments with model artifacts and bounded benchmark claims.

Open node

CIG

Persistent provenance-aware claim/evidence graph for bounded research claims.

Open node

Proof Ranker

Verifier-backed proof scoring and repair ladder published as Hugging Face artifacts.

Open node