GCTS MVP Build

The GCTS MVP can be built without full custom model training. The first target is an auditable decision-support prototype that accepts a bounded corpus, extracts evidence and claims, models access states, enumerates worlds, and emits ranked reports.

Product Boundary

The MVP should be an Evidence Accountability Workbench focused on auditable evidence operations. The first useful product should help analysts organize evidence, identify record contingencies, preserve contradiction, and report what records would change the analysis.

Initial users:

• investigative researchers;
• legal support teams;
• compliance analysts;
• journalists handling incomplete records;
• internal auditors;
• intelligence-style analytic teams.

Phase 1: Local Prototype

Use existing models and explicit schemas:

• LLMs for candidate extraction, latent-context suggestions, access-hypothesis suggestions, and rendering.
• Retrieval plus citation validation for evidence grounding.
• NLI or entailment models for claim-evidence scoring.
• A small evidence-access model for expected record existence, availability, control, and non-production.
• A rule compiler for a monotone tensor-logic core.
• Candidate-world enumeration or beam search.
• Calibration data to map evidence, access, incentive, and contradiction signals to probabilities.
• A dashboard to expose world rankings, proof traces, record-access states, uncertainty, and next evidence.

Fine-tuning is optional in Phase 1. If used, it should target extraction, evidence linking, access-state classification, and calibration. Direct runtime truth judgment stays outside model generation.

Runtime Data Products

The MVP should persist:

• evidence atoms;
• record-access states;
• institutional incentive profiles;
• claims and relations;
• rules and proof traces;
• world views;
• posterior and confidence reports;
• rendered synthesis reports.

API Surface

The first API should expose:

• POST /runs to create a synthesis run from a corpus manifest;
• GET /runs/{id} for run status;
• GET /runs/{id}/evidence for evidence atoms;
• GET /runs/{id}/access for record-access states;
• GET /runs/{id}/worlds for top-K possible worlds;
• GET /runs/{id}/claims for claim rankings and statuses;
• GET /runs/{id}/report for the rendered report.

MVP Gates

The first build succeeds only if it reaches:

• 100% resolvable citations for promoted strict claims;
• zero promoted zero-temperature claims without proof traces;
• calibrated claim probabilities with ECE at or below 0.10 on held-out verification tasks;
• top-3 world coverage at or above 85% on synthetic latent-context tasks;
• measurable chirality correlation with synthesis difficulty;
• measurable access-state calibration on adversarial record-suppression tasks;
• explicit distinction between unsupported, record_contingent, conflicted, and rejected;
• ablation evidence that multiverse/proof/access scoring beats simple RAG and LLM debate baselines on grounding, uncertainty quality, and likely-truth ranking.

First Repository Shape

gcts-prototype/
  gcts/
    schemas.py
    access_states.py
    rules.py
    worlds.py
    scoring.py
    statuses.py
    audit.py
  examples/
    facility_incident/
      evidence.json
      records.json
      claims.json
  outputs/
  README.md

First Demonstration

The first demo should show the same evidence under different access states:

1. Available record.
2. Inaccessible record.
3. Withheld record.
4. Not-generated record.
5. Evidence of absence.

The expected result is a visible status difference across runs, with strict proof, likely-truth posterior, and confidence reported separately.