{
  "@context": {
    "@language": "en",
    "@vocab": "https://schema.org/",
    "citeAs": "cr:citeAs",
    "column": "cr:column",
    "conformsTo": "dct:conformsTo",
    "cr": "http://mlcommons.org/croissant/",
    "data": {
      "@id": "cr:data",
      "@type": "@json"
    },
    "dataType": {
      "@id": "cr:dataType",
      "@type": "@vocab"
    },
    "dct": "http://purl.org/dc/terms/",
    "equivalentProperty": "cr:equivalentProperty",
    "examples": "cr:examples",
    "extract": "cr:extract",
    "field": "cr:field",
    "fileObject": "cr:fileObject",
    "fileProperty": "cr:fileProperty",
    "fileSet": "cr:fileSet",
    "format": "cr:format",
    "includes": "cr:includes",
    "isLiveDataset": "cr:isLiveDataset",
    "jsonPath": "cr:jsonPath",
    "key": "cr:key",
    "md5": "cr:md5",
    "parentField": "cr:parentField",
    "path": "cr:path",
    "prov": "http://www.w3.org/ns/prov#",
    "rai": "http://mlcommons.org/croissant/RAI/",
    "recordSet": "cr:recordSet",
    "references": "cr:references",
    "regex": "cr:regex",
    "repeated": "cr:repeated",
    "replace": "cr:replace",
    "samplingRate": "cr:samplingRate",
    "sc": "https://schema.org/",
    "separator": "cr:separator",
    "source": "cr:source",
    "subField": "cr:subField",
    "transform": "cr:transform"
  },
  "@type": "sc:Dataset",
  "name": "RAIL",
  "description": "RAIL is a CHC-grounded benchmark for large audio-language models with 5,306 items over 5 core capabilities (Auditory Processing, Reasoning, Memory, Processing Efficiency, Knowledge) and 32 sub-capabilities, covering speech, environmental sound and music.",
  "url": "https://huggingface.co/datasets/AIMS-RAIL/RAIL",
  "license": "https://spdx.org/licenses/CC-BY-4.0.html",
  "version": "2.0.0",
  "datePublished": "2026-06-09",
  "citeAs": "RAIL: Rethinking Auditory Intelligence in Large Audio-Language Models with a CHC-Grounded Benchmark. arXiv:2606.11260, 2026.",
  "conformsTo": "http://mlcommons.org/croissant/1.1",
  "distribution": [
    {
      "@type": "cr:FileSet",
      "@id": "manifests",
      "name": "manifests",
      "description": "Per-task JSONL manifests, one JSON object per item.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/tree/main/data",
      "encodingFormat": "application/jsonlines",
      "includes": "data/*/*.jsonl"
    },
    {
      "@type": "cr:FileSet",
      "@id": "audio_files",
      "name": "audio",
      "description": "Audio files referenced by relative paths in the manifests.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/tree/main/audio",
      "encodingFormat": [
        "audio/wav",
        "audio/mpeg",
        "audio/flac"
      ],
      "includes": "audio/**/*"
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_US",
      "name": "US.jsonl",
      "description": "Speech Sound Discrimination (US), Ga.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/auditory_processing/US.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "36672c476dd94df4c7ccf3ef57b953e89e456c2ab8bfa7a4d81377081c9525f2",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_UR",
      "name": "UR.jsonl",
      "description": "Resistance to Auditory Stimulus Distortion (UR), Ga.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/auditory_processing/UR.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "a90b460ef2a57ec406f0c8d273e884628d9c340037f7032430532f934f4d8219",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_U8",
      "name": "U8.jsonl",
      "description": "Maintaining and Judging Rhythm (U8), Ga.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/auditory_processing/U8.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "e934afeb5b07fdeaa017d83a9dfe8356d468115bd8f19064b6cc034c99c278f6",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_UP",
      "name": "UP.jsonl",
      "description": "Absolute Pitch (UP), Ga.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/auditory_processing/UP.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "3a0003459577dddae79b362141552528fa4075c34a2866a38497042ab1b6dd04",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_U1_U9",
      "name": "U1_U9.jsonl",
      "description": "Musical Discrimination and Judgment (U1/U9), Ga.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/auditory_processing/U1_U9.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "66b7fcba44fdecf0c33b7ea4ec1d42e5880ee7accd136c7fb493959dc37b751b",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_UL",
      "name": "UL.jsonl",
      "description": "Sound Localization (UL), Ga.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/auditory_processing/UL.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "f69f08021094f877174d84c36e5a94a6824e510abee207ba6d99e1f5948010e5",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_PC",
      "name": "PC.jsonl",
      "description": "Phonetic Coding (PC), Ga.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/auditory_processing/PC.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "c1d450bc13cb072e7c1047add0a8c693580cbc2423bd51097494a80cf4933858",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_I",
      "name": "I.jsonl",
      "description": "Induction (I), Gf.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/reasoning/I.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "bc8e2b1314be2c6e9efc4c1ce10a1f41c44dbaf0cfcb68f96382cfcf7ad43e01",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_RG",
      "name": "RG.jsonl",
      "description": "General Sequential Reasoning (RG), Gf.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/reasoning/RG.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "e92c6dd466c5e166787243b4b4abbbd9d3c862d36812961d24ea7978e4dbb8dc",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_RQ",
      "name": "RQ.jsonl",
      "description": "Quantitative Reasoning (RQ), Gf.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/reasoning/RQ.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "d07eedc75c1ce8b774f053d0471fbf239e614e50a81a289b6ec505fd7e0ff419",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_MS",
      "name": "MS.jsonl",
      "description": "Memory Span (MS), Gsm/Glr.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/memory/MS.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "c66f2fb6c8b0cc28f996a35c26319aa33719c5b15c3e6a5d2a01d3c019f12f94",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_MA",
      "name": "MA.jsonl",
      "description": "Associative Memory (MA), Gsm/Glr.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/memory/MA.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "e3286a4c318a016ada54732e3413c6bd46c0189bf9cd2fe7a84a1269a89a3ad3",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_MM",
      "name": "MM.jsonl",
      "description": "Meaningful Memory (MM), Gsm/Glr.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/memory/MM.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "139b7a7040b6dca7be81bbc2dd1562dccc608fe2ed561dbcd0fbbf5f04f0df84",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_M6",
      "name": "M6.jsonl",
      "description": "Free Recall Memory (M6), Gsm/Glr.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/memory/M6.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "715392511411cc8bd1f05898485736776251e764ab2fa05e5ee6be5cff1c2fe9",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_UM",
      "name": "UM.jsonl",
      "description": "Memory for Sound Patterns (UM), Gsm/Glr.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/memory/UM.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "8382d0b53948bc817b7da62b48a4779ad1f62f677dd3fbea7435efd71287e05f",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_WM",
      "name": "WM.jsonl",
      "description": "Working Memory (WM), Gsm/Glr.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/memory/WM.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "6ab5f72f9d11aeb2dea367b9d18cc0f35c49991098ce95389ca0d58b1b810684",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_P",
      "name": "P.jsonl",
      "description": "Perceptual Speed (P), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/P.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "ec66556d9feacc2f17dd1a01b9993b36a82f327d070d1a716e78d579c7ecd900",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_R9",
      "name": "R9.jsonl",
      "description": "Rate-of-Test-Taking (R9), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/R9.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "1e6c1c20a30da385e45c99f1bbe4807bac1c6a558c16ea059218cd2ff1bc6e3e",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_N",
      "name": "N.jsonl",
      "description": "Number Facility (N), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/N.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "fe785856696f53e9728e1f2288f87dba2db9e3c9fabcb21830e8280351f462ac",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_RS",
      "name": "RS.jsonl",
      "description": "Reading Speed (RS), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/RS.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "abcbb9f42e7020af9f96b61a7fb702d90b1c34fb31dcf84f640a31c2e87f1faf",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_R1",
      "name": "R1.jsonl",
      "description": "Simple Reaction Time (R1), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/R1.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "475e7c394f98ac0b2b986cbc370bec6463c63e422309cf125663d244220da1ca",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_R2",
      "name": "R2.jsonl",
      "description": "Choice Reaction Time (R2), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/R2.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "ab08a1d0d0228cfdb6a4b1dce9f756fd5711777044bf6162ae7ccb2cc1903478",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_R4",
      "name": "R4.jsonl",
      "description": "Semantic Processing Speed (R4), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/R4.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "07ea48ecc1b8126f0031a99d2dc0074572fcdee5af272d29083ea7a2bb616ce6",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_R7",
      "name": "R7.jsonl",
      "description": "Mental Comparison Speed (R7), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/R7.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "29f64ce345cbb4753381d026b66b2470657254467e666698e8d9aea53a032287",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_IT",
      "name": "IT.jsonl",
      "description": "Inspection Time (IT), Gs.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/processing_efficiency/IT.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "dce76d409b4817f8de75ca31df5145f2863ef6467ebc628640f065be6178fb11",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_K0",
      "name": "K0.jsonl",
      "description": "General (Verbal) Information (K0), Gc/Gkn.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/knowledge/K0.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "3eaded15c7481011438cf54ccd1efbe83fbcb4b8feba0ac58ec4ddf10027df8d",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_LD",
      "name": "LD.jsonl",
      "description": "Language Development (LD), Gc/Gkn.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/knowledge/LD.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "e72eefa75f08cd6c5ef7f63c066e5a11e61aa7aef387fabc1ff0b9d96007bec9",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_LS",
      "name": "LS.jsonl",
      "description": "Listening Ability (LS), Gc/Gkn.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/knowledge/LS.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "df261b26a2012177c2a98ae1ae6d5c67ddc3e27726cc1f71a7fac077dd5e1459",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_KL",
      "name": "KL.jsonl",
      "description": "Foreign Language Proficiency (KL), Gc/Gkn.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/knowledge/KL.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "e9b75870bd55aa3f5f47a2fb4e46afbf089f2b9eb9494d4e41a08a8e9159cb0d",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_A5",
      "name": "A5.jsonl",
      "description": "Geography Achievement (A5), Gc/Gkn.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/knowledge/A5.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "530182326f26e9762e298dc495e0a30e29b8b06f0f09c6211f0496ce325d600f",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_MK",
      "name": "MK.jsonl",
      "description": "Mechanical Knowledge (MK), Gc/Gkn.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/knowledge/MK.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "81214d980c70b93b3c007c55b2a8d94d9363e24aeeec476bb01fa2ea38995f0d",
      "containedIn": {
        "@id": "manifests"
      }
    },
    {
      "@type": "cr:FileObject",
      "@id": "task_BC",
      "name": "BC.jsonl",
      "description": "Knowledge of Behavioral Content (BC), Gc/Gkn.",
      "contentUrl": "https://huggingface.co/datasets/AIMS-RAIL/RAIL/resolve/main/data/knowledge/BC.jsonl",
      "encodingFormat": "application/jsonlines",
      "sha256": "703af444bf0f6631fbcf051f0040932d9f152d8b9e9595f0805158d66075afbb",
      "containedIn": {
        "@id": "manifests"
      }
    }
  ],
  "recordSet": [
    {
      "@type": "cr:RecordSet",
      "@id": "rail_records",
      "name": "rail_records",
      "description": "Item-level records of RAIL.",
      "field": [
        {
          "@type": "cr:Field",
          "@id": "rail_records/id",
          "name": "id",
          "description": "Unique item id.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "id"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/capability",
          "name": "capability",
          "description": "CHC core capability.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "capability"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/capability_code",
          "name": "capability_code",
          "description": "CHC broad-ability code.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "capability_code"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/task",
          "name": "task",
          "description": "Sub-capability (CHC narrow ability).",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "task"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/task_code",
          "name": "task_code",
          "description": "CHC narrow-ability code.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "task_code"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/subtask",
          "name": "subtask",
          "description": "Task design within the sub-capability, if any.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "subtask"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/answer_format",
          "name": "answer_format",
          "description": "multiple_choice, closed_label, open_ended or free_recall.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "answer_format"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/question",
          "name": "question",
          "description": "Question stem.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "question"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/choices",
          "name": "choices",
          "description": "Option texts or allowed output labels.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "choices"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/answer",
          "name": "answer",
          "description": "Gold answer text.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "answer"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/answer_label",
          "name": "answer_label",
          "description": "Gold option letter for multiple-choice items.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "answer_label"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/prompt",
          "name": "prompt",
          "description": "Full model-facing prompt used in evaluation.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "prompt"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/audio",
          "name": "audio",
          "description": "Primary relative audio path.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "audio"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/audio_paths",
          "name": "audio_paths",
          "description": "All relative audio paths in presentation order.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "audio_paths"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/num_audio",
          "name": "num_audio",
          "description": "Number of audio clips.",
          "dataType": "sc:Integer",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "num_audio"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/duration_s",
          "name": "duration_s",
          "description": "Total audio duration in seconds.",
          "dataType": "sc:Float",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "duration_s"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/source_id",
          "name": "source_id",
          "description": "Item id in the source manifest.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "source_id"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/legacy_id",
          "name": "legacy_id",
          "description": "Item id in RAIL v1.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "legacy_id"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/legacy_config",
          "name": "legacy_config",
          "description": "Config name in RAIL v1.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "legacy_config"
            }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rail_records/metadata_json",
          "name": "metadata_json",
          "description": "Original source row as JSON text.",
          "dataType": "sc:Text",
          "source": {
            "fileSet": {
              "@id": "manifests"
            },
            "extract": {
              "column": "metadata_json"
            }
          }
        }
      ]
    }
  ],
  "rai:dataBiases": "Potential biases include source-dataset selection bias, English-centric prompts, uneven coverage of speakers, environments, tasks, audio qualities, and model-facing question styles. Synthetic or templated tasks may over-represent clean, well-formed instructions compared with naturally occurring audio interactions.",
  "rai:dataLimitations": "The benchmark is intended for evaluating audio and auditory-cognitive reasoning, not for training production systems or making decisions about individuals. Coverage varies across source datasets, tasks, languages, speakers, environments, and synthetic generation procedures; results may not generalize to all accents, recording conditions, cultures, clinical settings, or real-world auditory tasks. Some rows aggregate processed versions of external datasets and should be used within the license and consent constraints of those sources.",
  "rai:dataSocialImpact": "Positive impact includes improving transparency and reproducibility for audio reasoning evaluation. Risks include overclaiming general audio understanding, amplifying source-dataset biases, or using speech attributes for profiling. Mitigations include preserving source metadata, documenting limitations, and recommending evaluation-only use.",
  "rai:dataUseCases": "Validated use cases are benchmark evaluation and diagnostic analysis of audio reasoning systems across the included task categories. The dataset is not validated for model training, human-subject assessment, medical or accessibility diagnosis, surveillance, biometric identification, or high-stakes decision making.",
  "rai:hasSyntheticData": true,
  "rai:personalSensitiveInformation": "Some source audio may contain human speech and therefore may encode or reveal sensitive attributes such as language, accent, perceived gender, age, emotion, or cultural background. The benchmark should not be used to identify, profile, or make consequential inferences about individuals.",
  "prov:wasDerivedFrom": [
    "https://huggingface.co/datasets/AIMS-RAIL/RAIL (v1.0.0)"
  ],
  "prov:wasGeneratedBy": [
    {
      "@type": "prov:Activity",
      "name": "RAIL v2 re-organisation",
      "description": "v1 rows were relabelled with the paper's capability/task/subtask taxonomy, given unique ids, normalised to a single answer schema, and memory prompts were rebuilt as used in evaluation."
    }
  ]
}