{
  "@context": "https://schema.org",
  "@type": "TechArticle",
  "@id": "https://anchorfact.org/kb/post-training-alignment",
  "headline": "Post-Training Alignment: DPO, GRPO, and Beyond",
  "description": "Post-training alignment makes base models useful and safe. DPO simplified RLHF by removing the reward model; GRPO enabled DeepSeek-R1's reasoning breakthrough. The alignment pipeline (SFT→DPO/RL→reasoning) is now standard across the industry.",
  "dateCreated": "2026-05-24T02:49:13.650Z",
  "dateModified": "2026-05-24",
  "author": {
    "@type": "Organization",
    "name": "AnchorFact"
  },
  "publisher": {
    "@type": "Organization",
    "name": "AnchorFact",
    "url": "https://anchorfact.org"
  },
  "license": "https://creativecommons.org/licenses/by/4.0/",
  "anchorfact:confidence": "high",
  "anchorfact:generationMethod": "ai_assisted",
  "citation": [
    {
      "@type": "CreativeWork",
      "name": "Direct Preference Optimization: Your Language Model is Secretly a Reward Model",
      "sameAs": "https://arxiv.org/abs/2305.18290"
    },
    {
      "@type": "CreativeWork",
      "name": "DeepSeek-R1 Technical Report",
      "sameAs": "https://arxiv.org/abs/2501.12948"
    }
  ]
}