Public Release - 2026-06-03

Every uploaded paper needs an attack queue.

The public value of the 16-paper upload matrix is not immunity from critique. Each claim needs a bounded route for counterexamples, stronger baselines, reproduction gaps, and boundary updates.

paper_id target_claim evidence_gap boundary_effect

Counterexample Queue Template Evidence Map Registries

Schema visual mapping generic challenges into paper_id, target_claim, evidence_gap, and boundary_effect fields.

Upload Matrix

Critique must be routed to a paper and a claim.

The current public record is tied to a 16-paper upload matrix: W0, P23, P28, P29, P30, P31, P32-P40, and F1/P8. The point is not paper-count promotion. The point is to make each paper's claim, evidence, boundary, freshness status, and attack route inspectable.

Today's artifact is a per-paper Counterexample Queue. A useful attack should name the paper, the target claim, the expected public evidence, the observed gap, and the boundary effect. It should not request private runtime, customer data, account state, internal logs, private prompts, or execution chains.

open

The counterexample names a paper_id, target claim, and public artifact route.

triaged

The report is assigned to an attack class: proof gap, replay gap, baseline, boundary, or authority leak.

reproduced

The observed gap can be checked with public material or a public-safe minimal fixture.

accepted

The claim must narrow, split, downgrade, request more evidence, or enter a no-go report.

rejected

The attack fails for public reasons, without relying on private material as the rebuttal.

closed

The boundary ledger records the public note, revision, or reason for closure.

Queue Map

Attack classes for the paper matrix.

Paper groupUseful attackPublic boundaryLedger effect

W0Metrology or interoperability claim gap.Public manifest and theorem evidence only.Adjust claim scope or evidence tier.

P23Dry-run to real self-modification gap.No mutation scripts, private prompts, or production runtime.Keep claim at bounded replay until stronger evidence exists.

P28 / P29Drift bridge, conflict family, or minority relabel failure.No customer traces or sensitive group data.Refine heldout replay, synthetic slice, or value-boundary note.

P30 / P31Proof transcript or honesty-bound overclaim.No complete ZK system or permanent honesty guarantee is implied.Downgrade to protocol-stage where needed.

P32-P40Protocol gate, ethics gate, spectral negative-control, or authority leak.No production handoff, legal advice, governance authority, or deployment proof.Update authority and diagnostic boundaries.

F1/P8No-trade refusal or no-alpha boundary failure.No trading strategy, account state, broker state, or alpha claim.Preserve negative-result framing.

Boundary filter visual separating public claims and artifacts from private logs, customer data, and execution chains.

Submission

A useful counterexample is small and public-safe.

The queue should not reward vague disbelief. It should reward exact target claims, minimal public fixtures, concrete evidence gaps, stronger baselines within the same boundary, and explicit proposed boundary effects.

The private boundary is part of the protocol. A challenge that requires credentials, private logs, customer data, account state, internal prompts, or execution chains is not a public counterexample.

Template

Minimum fields for a repairable attack.

FieldPurposePublic-safe ruleBoundary effect

paper_idNames the paper or group being challenged.Use the public matrix identifier.Routes the report.

target_claimIdentifies the exact claim under attack.No moving target or paraphrase-only report.Enables claim ledger update.

expected_evidenceStates what public evidence should support the claim.Do not request protected material.Reveals evidence gaps.

observed_gapDescribes the missing proof, failed replay, or stronger baseline.Use public artifacts or minimal fixtures.Triggers triage.

boundary_effectProposes shrink, split, downgrade, withdraw, or next evidence.Keep the result public and bounded.Updates the ledger.

Challenge

Attack the public claim, not the private system.

Submit the smallest public-safe case that changes a boundary. The strongest critique is the one that leaves a repairable record: a narrower claim, a clearer evidence requirement, a stronger baseline, or a no-go note.

The standard remains: public credibility, not public control authority.