{
  "version": "1.0",
  "type": "link",
  "title": "Multi-Vendor Paired Evals (Anthropic, OpenAI, Google) & Basel LEX Caps | MARKOWITZ.SYSTEMS",
  "description": "Measure empirical cross-vendor error correlation ρ across Anthropic, OpenAI, and Google via paired evals and enforce Basel Large-Exposure concentration caps per foundation model.",
  "author_name": "MARKOWITZ.SYSTEMS Quantitative Finance & Continuing Professional Education Collective",
  "author_url": "https://markowitz.systems/education",
  "provider_name": "MARKOWITZ.SYSTEMS — Fresh AI Preprints & Professional Upskilling",
  "provider_url": "https://markowitz.systems",
  "cache_age": 86400,
  "thumbnail_url": "https://markowitz.systems/og-card.png",
  "thumbnail_width": 1200,
  "thumbnail_height": 630,
  "url": "https://markowitz.systems/vendor-diversification-evals",
  "fresh_preprints_url": "https://markowitz.systems/preprints",
  "continuing_education_url": "https://markowitz.systems/education",
  "meta_education_url": "https://markowitz.systems/meta-education"
}