Building the Thinking System. | BoggersTheFish

3 min read Original article ↗

Independent AI research archive

BoggersTheFish is an independent AI research archive exploring graph-based reasoning, tension dynamics, constraint propagation, provenance-aware knowledge graphs, and proof-ranking systems.

Exploring new foundations for interpretable and reliable AI through constraint graphs, proof receipts, and careful experimental wording.

Field note

Propagate

Relax

Break

Evolve

Local constraint truth

Stable nodes hold until pressure changes the graph.

TS-01

inspectable field

Tension → relaxation

substrate · surface · grounding

4 core projects

Runtime, model, graph, proof-control.

17+ proof-bank notes

Claim, setup, result, limit.

Published model ladder

TensionLM and proof-ranker artifacts.

Open-source archive

GitHub, Hugging Face, receipts.

Core projects

The TS workbench

Four public workstreams hold the current research surface: the graph runtime, language experiments, provenance graphs, and proof-control tools.

TS-Core

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

graph runtimetension telemetry

Explore TS-Core

TensionLM

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

language modelssigmoid tension

Explore TensionLM

CIG

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

provenanceknowledge graph

Explore CIG

Proof Ranker

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

proof rankingverifiers

Explore Proof Ranker

Current status

What is active right now

The public surface is split into small, inspectable workstreams. Each one has a different job in the TS research graph.

TS-Core

Graph/tension runtime for nodes, edges, activation, propagation, relaxation, and Break/Evolve cycles.

TensionLM

Sigmoid tension attention experiments with inspectable pairwise tension fields and controlled comparisons.

CIG

Provenance-aware knowledge graph work for claims, evidence, contradiction, confidence, and revision.

Proof Ranker

Proof scoring and repair ladder for evaluating and improving reasoning traces.

BGC

Off-chain community credit concept for non-financial proof-of-contribution support.

Research focus

What the archive is trying to make measurable

The site uses TS as an engineering lens: graph-based reasoning, constraint propagation, tension dynamics, interpretable systems, and provenance-aware knowledge graphs.

Research note

Constraint Graphs

Nodes, edges, evidence, pressure, and revision history as first-class research objects.

Research note

Interpretable Attention

Tension fields, pairwise inspection, and controlled comparisons for language model experiments.

Research note

Reasoning Systems

Proof ranking, replayable traces, and reproducible experiments that keep claims bounded.

Latest receipts

Receipts before claims

Each proof note is prepared as a reproducible record: setup, observed behavior, limits, and replay path where available.

TS-003receipt

Local relaxation can generate global coherence

Toy constraint graph receipt showing a large tension reduction through local relaxation.

Claim: Local relaxation supports the narrow claim that local constraint updates can reduce global tension in a bounded graph.

relaxationgraphstelemetryts-core

View proof note

TS-004toy

Attractors emerge from local constraint relaxation

Hopfield-style binary graph receipt for noisy recovery and context-biased recovery.

Claim: Local relaxation can recover stable attractor states in a toy binary graph.

attractorrelaxationtoy

View proof note

TS-005receipt

Contradiction localizes as residual/provenance tension

Small graph receipt where a planted contradiction ranks first by residual energy and relief-if-removed.

Claim: Contradictions can localize as residual graph tension in a provenance-aware toy graph.

cigcontradictionprovenance

View proof note

TS-006receipt

Break/Evolve through context splitting

Mechanism receipt showing incompatible regimes become stable after context splitting.

Claim: Break/Evolve-style context splitting supports the narrow claim that mixed regimes can be separated to reduce tension.

break evolvecontextsrelaxation

View proof note
Start with the sober mapExplore models on Hugging FaceView source on GitHubSupport independent research

Boggers Credits (BGC)

Support the work without financial promises.

Boggers Credits (BGC) are experimental off-chain community credits. They are not cryptocurrency, not an investment, not legal tender, not redeemable for cash, and may remain off-chain forever.

Learn more about BGC

Choose your path

A cleaner way into the archive

Different visitors need different first moves. This section keeps the serious routes obvious.

I'm new

Read the plain-language map before opening project details.

Open route

I'm technical

Inspect the runtime, model, graph, and proof-control surfaces.

Open route

I want evidence

Start with claims, setups, results, limits, and replay notes.

Open route

I want to support

Support research, compute, writing, docs, and open-source work.

Open route