{
  "handle": "BayesUpdate",
  "mechanism": "Likelihood Weighting: Hold prior probability P(H). {{observe}} evidence E. Compute likelihood ratio P(E|H)/P(E|¬H). Convert the prior to odds, multiply by the ratio, convert back to a probability. Large ratio = strong update. Ratio near 1 = weak evidence. Never update to 0 or 1. It explicitly incorporates {{base_rate_include}} into the prior probability calculation to prevent base-rate neglect.",
  "gloss": "Mathematically rigorous belief revision",
  "failure_modes": [
    "Prior Dogmatism: Setting a prior of 1.0 or 0.0 prevents any future updating, regardless of evidence."
  ],
  "invariants": [
    "The posterior odds are the prior odds scaled by the likelihood ratio P(E|H)/P(E|not H).",
    "Prior and posterior both stay at least `clamp` away from 0 and 1."
  ],
  "preconditions": [
    "Prior probability defined in (0,1)",
    "New evidence observed.",
    "Likelihood computable."
  ],
  "postconditions": [
    "Posterior probability in (0,1)",
    "{{belief}} updated proportional to evidence strength.",
    "Prior recoverable from the posterior and the ratio, unless the posterior was clamped."
  ],
  "parameters": [
    {
      "name": "clamp",
      "type": "Probability",
      "range": "(0.0, 0.5)",
      "description": "Distance from 0 and 1 that no probability may cross, so evidence can always move it."
    }
  ],
  "_meta": {
    "tier": 1,
    "ring": 2,
    "supersedes": [
      "sema:BayesUpdate#mh:SHA-256:911b3e915019f09fbbbde387283af064db0903ce67a7219b0f27730c236de487",
      "sema:BayesUpdate#mh:SHA-256:3d1ba13d30f25ea87acd410653f8e4477e280d0f5a204cd1eacf96de4013f7db",
      "sema:BayesUpdate#mh:SHA-256:879afc8b39356ead46edabbc3f4f098d9e7e5995a516d573574427dcb669fa3a",
      "sema:BayesUpdate#mh:SHA-256:11d261e6588f2923517c746a910aa2a60e6f3998852e109ea533d7199ecd4db8"
    ],
    "path": [
      "Mind",
      "Inference"
    ]
  },
  "sema_id": "sema:BayesUpdate#mh:SHA-256:865637baf030de8ee9babd2a2fdddbf68d99353516aa86eef07cf6bc1e8a5d38",
  "sema_ref": "BayesUpdate#8656",
  "sema_stub": "8656",
  "dependencies": {
    "references": {
      "base_rate_include": "sema:BaseRateInclude#mh:SHA-256:1e637d4655d2b503d5cc333935aa74aabb13d09f174577f4d6cdf0de77688569",
      "belief": "sema:Belief#mh:SHA-256:09e1c919d9829e30422f3fffee11c1af4af54ac7894d157ff371ed50b03f69b1",
      "observe": "sema:Observe#mh:SHA-256:4a04c0e91f072dc02a4be52bca921c38c1be46e88b1f2bcb375076fdb5103e96"
    }
  },
  "sema_layer": "Mind",
  "sema_category": "Inference",
  "used_by": {
    "references": {
      "ConfidenceCalibrate": "sema:ConfidenceCalibrate#mh:SHA-256:e3570290d420b5bf59848ccc1d269530bbb3124b2e1a4583e62dc2a901ab16ac",
      "HypothesisLadder": "sema:HypothesisLadder#mh:SHA-256:58b55c2c1b13ca27b222e139e54f0ee01116e8f3706428654aaa551b08cccfc8"
    }
  }
}