outputs/claim1.json

{
  "all_formula_rows_match": true,
  "all_small_enumerations_match": true,
  "assessment": "verified",
  "claim": 1,
  "destructive_control": {
    "b": 2,
    "correct": 4200,
    "d": 4,
    "m": 3,
    "without_nonempty_train_test_exclusion": 6804
  },
  "destructive_control_detected": true,
  "literal_claim": "DiscoGen procedurally generates over 400 million distinct algorithm discovery tasks via a combinatorial formula N_tasks = 2*3*b*(2^m-1)*(3^d-2^(d+1)+1) depending on the number of modules m, datasets d, and backends b (Section 4.2, Equation 1).",
  "main_domain_rows": [
    {
      "b": 1,
      "d": 11,
      "domain": "Bayesian Optimisation",
      "m": 6,
      "matches": true,
      "recomputed_tasks": 65413656,
      "reported_tasks": 65413656
    },
    {
      "b": 1,
      "d": 7,
      "domain": "Brain Speech Detection",
      "m": 3,
      "matches": true,
      "recomputed_tasks": 81144,
      "reported_tasks": 81144
    },
    {
      "b": 1,
      "d": 9,
      "domain": "Computer Vision Classification",
      "m": 4,
      "matches": true,
      "recomputed_tasks": 1679400,
      "reported_tasks": 1679400
    },
    {
      "b": 3,
      "d": 3,
      "domain": "Continual Learning",
      "m": 5,
      "matches": true,
      "recomputed_tasks": 6696,
      "reported_tasks": 6696
    },
    {
      "b": 1,
      "d": 4,
      "domain": "Greenhouse Gas Prediction",
      "m": 2,
      "matches": true,
      "recomputed_tasks": 900,
      "reported_tasks": 900
    },
    {
      "b": 2,
      "d": 4,
      "domain": "Language Modelling",
      "m": 3,
      "matches": true,
      "recomputed_tasks": 4200,
      "reported_tasks": 4200
    },
    {
      "b": 1,
      "d": 3,
      "domain": "Model Unlearning",
      "m": 1,
      "matches": true,
      "recomputed_tasks": 85176,
      "reported_tasks": 85176
    },
    {
      "b": 1,
      "d": 4,
      "domain": "Off-Policy RL",
      "m": 7,
      "matches": true,
      "recomputed_tasks": 38100,
      "reported_tasks": 38100
    },
    {
      "b": 3,
      "d": 13,
      "domain": "On-Policy RL",
      "m": 4,
      "matches": true,
      "recomputed_tasks": 426043800,
      "reported_tasks": 426043800
    },
    {
      "b": 1,
      "d": 4,
      "domain": "Unsupervised Environment Design",
      "m": 3,
      "matches": true,
      "recomputed_tasks": 2100,
      "reported_tasks": 2100
    }
  ],
  "main_total_from_rows": 493355172,
  "official_model_choices": 13,
  "over_400_million": true,
  "reported_main_total": 493355172,
  "small_exact_enumerations": [
    {
      "b": 1,
      "d": 2,
      "direct": 12,
      "formula": 12,
      "m": 1,
      "matches": true
    },
    {
      "b": 1,
      "d": 3,
      "direct": 216,
      "formula": 216,
      "m": 2,
      "matches": true
    },
    {
      "b": 2,
      "d": 4,
      "direct": 4200,
      "formula": 4200,
      "m": 3,
      "matches": true
    },
    {
      "b": 3,
      "d": 4,
      "direct": 13500,
      "formula": 13500,
      "m": 4,
      "matches": true
    }
  ]
}