Part of CNS 7.1 / GCTS: Grounded Chiral Tensor Synthesis

Adversarial Evidence And Access Modeling

GCTS is designed for environments where evidence is limited, contradictory, or adversarially curated. The central discipline is simple: absence has multiple states.

Five Absence States

  1. Absence of evidence: no available supporting evidence has been found.
  2. Evidence of absence: an expected record or observation exists and affirmatively negates the claim.
  3. Inaccessible evidence: the record may exist outside the runtime corpus.
  4. Withheld evidence: non-production is more likely under a withholding world than under a benign-missingness world.
  5. Not-generated evidence: the record should not be expected to exist.

Only evidence of absence can directly penalize a claim as absent. Other states usually create access uncertainty, record contingencies, or competing worlds.

Access Features

For each expected record, GCTS models:

  • who controls it;
  • whether ordinary procedure would generate it;
  • whether it should be observable;
  • whether it was requested, produced, refused, partially produced, contradicted, destroyed, sealed, or unavailable;
  • confidence in that access-state classification.

Incentive Features

Institutional incentive profiles model:

  • control over records or testimony;
  • reputational, legal, financial, operational, or political exposure;
  • incentive to disclose;
  • incentive to conceal, delay, narrow, or frame evidence;
  • expected penalty if concealment is detected;
  • prior source reliability.

Incentives affect missingness likelihood, source quality, and world energy while leaving proof to evidence and rules. Claims still require evidence and rules.

Suppression Discipline

The system should infer strategic withholding only when several conditions line up:

  • a record was expected to exist;
  • a responsible actor plausibly controlled it;
  • the access path was legitimate or ordinary;
  • non-production is less likely under benign missingness;
  • the hypothesis reduces contradiction or explains access asymmetry without excessive unsupported complexity.

Unsupported suppression hypotheses should increase parsimony penalty.

Output Requirements

Any report involving missing or controlled evidence should state:

  • which records matter;
  • expected generation duty;
  • observed access state;
  • confidence in that classification;
  • whether the claim is record_contingent;
  • what evidence would raise, lower, or resolve the claim ranking.
Step 6 of 7 in CNS 7.1 / GCTS: Grounded Chiral Tensor Synthesis