{
  "format_version": 1,
  "paper": "arXiv:2607.15161",
  "seed": 42,
  "endpoint": "https://hf-mirror.com",
  "total_rows": 100000,
  "domain_counts": {
    "code": 33333,
    "math": 33334,
    "science": 33333
  },
  "unique_prompt_hashes": 74644,
  "duplicate_prompt_rows": 25356,
  "sampling": {
    "method": "balanced allocation over randomly ordered Parquet shards, then random row groups and rows within each shard",
    "max_source_shards": 12,
    "all_source_shards": false,
    "deduplicate_prompts": false,
    "quality_filter": null,
    "length_filter": null,
    "answers_and_traces_retained": false
  },
  "sources": [
    {
      "domain": "math",
      "dataset": "nvidia/OpenMathReasoning",
      "config": "default",
      "split": "cot",
      "prompt_field": "problem",
      "id_field": null,
      "source_revision": "d3d08664755704f422af97d43a7ff0ded4bd95df",
      "available_parquet_shards": 144,
      "selected_parquet_shards": [
        28,
        6,
        70,
        62,
        57,
        35,
        26,
        139,
        22,
        108,
        8,
        7
      ],
      "selected_rows": 33334,
      "unique_source_ids": 30773,
      "duplicate_source_id_rows": 2561,
      "unique_prompt_hashes": 30773,
      "duplicate_prompt_rows": 2561,
      "prompt_characters": {
        "min": 18,
        "median": 168.0,
        "mean": 196.38285234295313,
        "max": 1594
      }
    },
    {
      "domain": "science",
      "dataset": "nvidia/OpenScienceReasoning-2",
      "config": "default",
      "split": "train",
      "prompt_field": "input",
      "id_field": null,
      "source_revision": "174b02c9cdf231f220765b2a1d5ece4550921894",
      "available_parquet_shards": 3,
      "selected_parquet_shards": [
        0,
        1,
        2
      ],
      "selected_rows": 33333,
      "unique_source_ids": 33333,
      "duplicate_source_id_rows": 0,
      "unique_prompt_hashes": 33333,
      "duplicate_prompt_rows": 0,
      "prompt_characters": {
        "min": 193,
        "median": 1107,
        "mean": 1154.49920499205,
        "max": 5179
      }
    },
    {
      "domain": "code",
      "dataset": "nvidia/OpenCodeReasoning",
      "config": "split_0",
      "split": "split_0",
      "prompt_field": "input",
      "id_field": "id",
      "source_revision": "20a1ca19c0d050fe9057fc08339d6b370ec1c67a",
      "available_parquet_shards": 9,
      "selected_parquet_shards": [
        6,
        1,
        7,
        3,
        8,
        2,
        0,
        5,
        4
      ],
      "selected_rows": 33333,
      "unique_source_ids": 10538,
      "duplicate_source_id_rows": 22795,
      "unique_prompt_hashes": 10538,
      "duplicate_prompt_rows": 22795,
      "prompt_characters": {
        "min": 28,
        "median": 1538,
        "mean": 1665.6388263882639,
        "max": 12973
      }
    }
  ]
}