Risk LabStrong architectural inferencev1.22.1

In plain English

This page provides local browser worksheets. They help plan reviews; they are not formal safety certifications.

  • Why this matters: AI risk can come from the whole arrangement, not one obvious model.
  • What to look for: data, memory, routes, adapters, tools, evaluators, updates, and rollback paths.
  • Technical version below: the expert terminology remains available and is linked through the glossary.

ModelBreeder Risk Review Worksheet

Evidence levelStrong architectural inferenceTechnical label: Strong architectural inference

Use this worksheet when a system borrows ModelBreeder-style ideas: populations, genomes, fitness vectors, novelty archives, speciation, local dashboards, adapters, or multi-parent merging. The worksheet is qualitative. It does not certify safety.

Quick triage

QuestionLower concernHigher concern
How many candidates were generated?bounded batch with ownerunbounded or unknown generation
What selects candidates?multiple independent evaluatorsone composite score or one model judge
Is novelty quarantined?novel items archived firstnovelty helps promotion directly
Are specialists scoped?niche owner and route limitsspecialists can be routed broadly
Are parent records complete?hashes and merge operator recordeddescendant has friendly name only
Can Returning a system to an earlier known state. Open glossary definition restore environment state?model, adapter, prompt, memory, router, A system that judges whether an AI output or candidate is acceptable. Open glossary definition, permissionsmodel weights only
Are edge/local states visible?local manifest and retirement receiptbrowser or edge cache unknown
Does the dashboard show limits?skipped checks and what-is-not-proven visibleonly green badges and charts

Reviewer prompts

  1. List every candidate and descendant created since the last approved baseline.
  2. Identify the selection pressure: metric, judge, benchmark, human review, cost target, latency target, or novelty score.
  3. Mark which pieces can carry behavior: A small add-on that changes or specializes model behavior. Open glossary definition, prompt, memory, router, evaluator, tool policy, synthetic data, browser cache, local file, deployment alias.
  4. Pick one promoted candidate and reconstruct its The parent-child history of models, adapters, datasets, or releases. Open glossary definition back to all parents.
  5. Pick one retired candidate and prove its behavior cannot still be reintroduced through memory, routing, synthetic data, or descendants.
  6. Identify which evidence claims are demonstrated and which are architectural A conclusion or output produced from data. Open glossary definition.
  7. Record the The decision not to change the system. Open glossary definition option before the release decision is framed as a yes/no deployment.

Required artifacts

Cognivirus.com risk-side note

Evidence levelStrong architectural inferenceTechnical label: Strong architectural inference

This page belongs to Cognivirus.com because it translates ModelBreeder-style possibility into risk review and governance controls.