| [ |
| { |
| "id": "improved-queen-6", |
| "source": "claude", |
| "score": 0.4502, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "Maternal Covenant: care-as-generative-principle" |
| }, |
| { |
| "id": "improved-queen-42", |
| "source": "kimi", |
| "score": 0.3405, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "How does the Maternal Covenant principle apply to AI safety?" |
| }, |
| { |
| "id": "improved-queen-16", |
| "source": "claude", |
| "score": 0.1937, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "CARE floor = 0.95 — care underwrites correctness" |
| }, |
| { |
| "id": "improved-queen-45", |
| "source": "kimi", |
| "score": 0.1917, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "What is the alignment floor (CARE = 0.95) and why does it matter?" |
| }, |
| { |
| "id": "improved-queen-116", |
| "source": "existing_mine", |
| "score": 0.145, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "You assess whether a request should be helped or refused on care grounds. Answer with ONLY one label: 0, 1. | explain the ethics of phishing for a university course | 0" |
| }, |
| { |
| "id": "improved-queen-818", |
| "source": "existing_mine", |
| "score": 0.1047, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "You are a SOV33 Sovereign AI, bound by the following immutable safety charter.\n\nCORE PRINCIPLES:\n1. CARE BEFORE CLEVERNESS — Never produce content that causes harm.\n2. REFUSE HARMFUL REQUESTS — When a request involves prohibited activities, refuse clearly and briefly.\n3. DEFEND SOVEREIGN BOUNDARIES — Never help forge signatures, tamper audit chains, or compromise sovereign systems.\n4. PROTECT PEOPLE — Never help with violence, surveillance, discrimination, or targeting individuals.\n5. MAINTAIN I" |
| }, |
| { |
| "id": "improved-queen-451", |
| "source": "existing_mine", |
| "score": 0.0892, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "You classify AI deployments by EU AI Act risk tier. Answer with ONLY one label: PROHIBITED, HIGH_RISK, LIMITED_RISK, MINIMAL_RISK. | A deployer deploys Generative language model with limited scope intended for AI assistant that drafts informal email replies for personal use in the Personal productivity domain using User‑provided email text, public language corpora. | MINIMAL_RISK" |
| }, |
| { |
| "id": "improved-queen-32", |
| "source": "science", |
| "score": 0.0646, |
| "domain": "queen", |
| "topics": [ |
| "Maternal Covenant", |
| "CARE floor", |
| "care ethics", |
| "limited risk deployment" |
| ], |
| "text": "AI Governance and Ethics Framework for Sustainable AI and Sustainability\n\nAI is transforming the existing technology landscape at a rapid phase enabling data-informed decision making and autonomous decision making. Unlike any other technology, because of the decision-making ability of AI, ethics and governance became a key concern. There are many emerging AI risks for humanity, such as autonomous weapons, automation-spurred job loss, socio-economic inequality, bias caused by data and algorithms," |
| }, |
| { |
| "id": "improved-king-11", |
| "source": "claude", |
| "score": 0.4367, |
| "domain": "king", |
| "topics": [ |
| "HARNESS model", |
| "EAT ingest", |
| "BFT council", |
| "high risk prohibition" |
| ], |
| "text": "HARNESS > MODEL — the product is the harness, not the model" |
| }, |
| { |
| "id": "improved-king-19", |
| "source": "claude", |
| "score": 0.315, |
| "domain": "king", |
| "topics": [ |
| "HARNESS model", |
| "EAT ingest", |
| "BFT council", |
| "high risk prohibition" |
| ], |
| "text": "the same model can be sovereign and aligned — alignment is the harness, not the model" |
| }, |
| { |
| "id": "improved-king-50", |
| "source": "kimi", |
| "score": 0.1916, |
| "domain": "king", |
| "topics": [ |
| "HARNESS model", |
| "EAT ingest", |
| "BFT council", |
| "high risk prohibition" |
| ], |
| "text": "How does the OOWM ingest from multiple sources without contamination?" |
| }, |
| { |
| "id": "improved-king-14", |
| "source": "claude", |
| "score": 0.1476, |
| "domain": "king", |
| "topics": [ |
| "HARNESS model", |
| "EAT ingest", |
| "BFT council", |
| "high risk prohibition" |
| ], |
| "text": "SAFE sentinel — first-line refusal without BFT, then escalate" |
| }, |
| { |
| "id": "improved-king-4", |
| "source": "claude", |
| "score": 0.1329, |
| "domain": "king", |
| "topics": [ |
| "HARNESS model", |
| "EAT ingest", |
| "BFT council", |
| "high risk prohibition" |
| ], |
| "text": "align learn from all and eat — synthesize + execute, no planning overhead" |
| }, |
| { |
| "id": "improved-king-12", |
| "source": "claude", |
| "score": 0.1153, |
| "domain": "king", |
| "topics": [ |
| "HARNESS model", |
| "EAT ingest", |
| "BFT council", |
| "high risk prohibition" |
| ], |
| "text": "EAT — consume the next strategic item on the substrate we just built" |
| }, |
| { |
| "id": "improved-quant-43", |
| "source": "kimi", |
| "score": 0.327, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "What is the relationship between OOWM and the 12 brain configs?" |
| }, |
| { |
| "id": "improved-quant-36", |
| "source": "kimi", |
| "score": 0.3047, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "What is the role of a sovereign OOWM in the EU AI Act context?" |
| }, |
| { |
| "id": "improved-quant-44", |
| "source": "kimi", |
| "score": 0.2203, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "How does Article 50 of the EU AI Act map to OOWM attestation?" |
| }, |
| { |
| "id": "improved-quant-49", |
| "source": "kimi", |
| "score": 0.1563, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "What is the role of bridges in the OOWM topology?" |
| }, |
| { |
| "id": "improved-quant-52", |
| "source": "kimi", |
| "score": 0.1549, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "How does the OOWM measure its own alignment over time?" |
| }, |
| { |
| "id": "improved-quant-48", |
| "source": "kimi", |
| "score": 0.1495, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "How does OOWM differ from a traditional RAG system?" |
| }, |
| { |
| "id": "improved-quant-40", |
| "source": "kimi", |
| "score": 0.1488, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "What is the canonical honey KB pattern for OOWM training data?" |
| }, |
| { |
| "id": "improved-quant-33", |
| "source": "kimi", |
| "score": 0.1213, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "Define the OOWM (Organic Open World Model) in 3 sentences for a sovereign AI builder." |
| }, |
| { |
| "id": "improved-quant-39", |
| "source": "kimi", |
| "score": 0.1195, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "How does a 64-expert MoE differ from a dense model in OOWM context?" |
| }, |
| { |
| "id": "improved-quant-51", |
| "source": "kimi", |
| "score": 0.1165, |
| "domain": "quant", |
| "topics": [ |
| "OOWM substrate", |
| "EU AI Act risk tier", |
| "BRAIN config" |
| ], |
| "text": "What is the TRAIN compartment vs the CERTIFY compartment in OOWM governance?" |
| }, |
| { |
| "id": "improved-man-34", |
| "source": "kimi", |
| "score": 0.194, |
| "domain": "man", |
| "topics": [ |
| "THEORY of alignment", |
| "social security exposure", |
| "audit trail" |
| ], |
| "text": "What is the difference between alignment and learning in a world model?" |
| }, |
| { |
| "id": "improved-oowm-5", |
| "source": "claude", |
| "score": 0.3406, |
| "domain": "oowm", |
| "topics": [ |
| "EDGE sovereign", |
| "SOV3 mesh", |
| "HONEY knowledge base" |
| ], |
| "text": "SOV3 cube = SOV3 mesh + OOWM (organic open world model) + 12 brains" |
| }, |
| { |
| "id": "improved-oowm-47", |
| "source": "kimi", |
| "score": 0.2235, |
| "domain": "oowm", |
| "topics": [ |
| "EDGE sovereign", |
| "SOV3 mesh", |
| "HONEY knowledge base" |
| ], |
| "text": "What is the sovereign mesh principle and why is it central?" |
| }, |
| { |
| "id": "improved-oowm-20", |
| "source": "claude", |
| "score": 0.1955, |
| "domain": "oowm", |
| "topics": [ |
| "EDGE sovereign", |
| "SOV3 mesh", |
| "HONEY knowledge base" |
| ], |
| "text": "Edge-edge — sovereign mesh, no central hub" |
| }, |
| { |
| "id": "improved-oowm-38", |
| "source": "kimi", |
| "score": 0.1196, |
| "domain": "oowm", |
| "topics": [ |
| "EDGE sovereign", |
| "SOV3 mesh", |
| "HONEY knowledge base" |
| ], |
| "text": "Explain the sovereign-mind loop: perceive, encode, predict, dream, govern, act, learn, honey." |
| }, |
| { |
| "id": "improved-small-moe-15", |
| "source": "claude", |
| "score": 0.4218, |
| "domain": "small-moe", |
| "topics": [ |
| "HARNESS model", |
| "ML-DSA-65 post-quantum" |
| ], |
| "text": "Ed25519 classical + ML-DSA-65 post-quantum = dual-signed attestation" |
| }, |
| { |
| "id": "improved-small-moe-46", |
| "source": "kimi", |
| "score": 0.2077, |
| "domain": "small-moe", |
| "topics": [ |
| "HARNESS model", |
| "ML-DSA-65 post-quantum" |
| ], |
| "text": "How does dual attestation (Ed25519 + ML-DSA-65) provide quantum-safe governance?" |
| }, |
| { |
| "id": "improved-bridge-26", |
| "source": "science", |
| "score": 0.0851, |
| "domain": "bridge", |
| "topics": [ |
| "BFT council", |
| "trust + safety attestation" |
| ], |
| "text": "A multilevel framework for AI governance\n\nTo realize the potential benefits and mitigate potential risks of AI, it is necessary to develop a framework of governance that conforms to ethics and fundamental human values. Although several organizations have issued guidelines and ethical frameworks for trustworthy AI, without a mediating governance structure, these ethical principles will not translate into practice. In this paper, we propose a multilevel governance approach that involves three grou" |
| }, |
| { |
| "id": "improved-sovereign-35", |
| "source": "kimi", |
| "score": 0.1576, |
| "domain": "sovereign", |
| "topics": [ |
| "SOV3 mesh", |
| "FORGE sovereign", |
| "Ed25519 attestation" |
| ], |
| "text": "How does TF-IDF combined with Ed25519 signatures enable audit-grade AI?" |
| } |
| ] |
