probe_json/counting_comparison/eval.json

[
  {
    "voice": [
      "counting/task_comparison/counting_pair_014.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two snippets are played with a brief pause between them. Pick 'A' if the first clip has more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_041.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two recordings are joined by a short silence. Answer only with a single letter: 'A' if the first segment sounds like it has more sound times, 'B' if the second segment sounds like it has more sound times. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_016.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two sound fragments are concatenated with a brief gap. Choose 'A' if the first has more sound occurrences, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_023.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You will hear two consecutive audio clips separated by silence. Select 'A' if the first has more sound events, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_042.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You are given two clips with a short silent gap in between. Say only 'A' if the first clip has more sound events, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_032.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A dual-segment audio is provided with a short pause. Only answer 'A' (first has more) or 'B' (second has more) based on sound count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_066.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two audio parts are separated by a moment of silence. Answer 'A' if the first has more sound occurrences, 'B' for the second. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_024.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two sound sequences are separated by a short silence. Indicate with 'A' if the first segment has more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_057.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two auditory segments are presented with a short gap. Your answer should be 'A' if the first has more sound instances, 'B' otherwise. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_037.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two auditory segments are presented with a short gap. Your answer should be 'A' if the first has more sound instances, 'B' otherwise. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_075.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two audio segments are joined by a moment of silence. Use 'A' if the first segment has more sounds, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_099.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "An audio contains two parts separated by a short silence. Answer 'A' if the first has more sound bursts, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_085.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You will hear two audio segments separated by a short silence. State only 'A' if the first has more sound pulses, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_036.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two sound fragments are concatenated with a brief gap. Choose 'A' if the first has more sound occurrences, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_052.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A dual-segment audio is provided with a short pause. Only answer 'A' (first has more) or 'B' (second has more) based on sound count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_004.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two sound sequences are separated by a short silence. Indicate with 'A' if the first segment has more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_009.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You are presented with two clips and a short silent interval. Only reply 'A' if the first segment contains more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_028.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A concatenated audio presents two segments with a brief pause. Return only 'A' (first has more) or 'B' (second has more) by perceived count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_069.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You are presented with two clips and a short silent interval. Only reply 'A' if the first segment contains more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_038.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Listen to two sound clips with a brief pause between them. Respond with just 'A' if the first has more sound occurrences, or 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_012.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A dual-segment audio is provided with a short pause. Only answer 'A' (first has more) or 'B' (second has more) based on sound count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_056.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two sound fragments are concatenated with a brief gap. Choose 'A' if the first has more sound occurrences, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_001.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two recordings are joined by a short silence. Answer only with a single letter: 'A' if the first segment sounds like it has more sound times, 'B' if the second segment sounds like it has more sound times. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_054.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two snippets are played with a brief pause between them. Pick 'A' if the first clip has more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_092.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A dual-segment audio is provided with a short pause. Only answer 'A' (first has more) or 'B' (second has more) based on sound count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_055.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two audio segments are joined by a moment of silence. Use 'A' if the first segment has more sounds, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_082.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You are given two clips with a short silent gap in between. Say only 'A' if the first clip has more sound events, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_039.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "An audio contains two parts separated by a short silence. Answer 'A' if the first has more sound bursts, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_019.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "An audio contains two parts separated by a short silence. Answer 'A' if the first has more sound bursts, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_000.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two tonal sequences are divided by a short silence. Enter 'A' if the first segment has more sound repetitions, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_045.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You will hear two audio segments separated by a short silence. State only 'A' if the first has more sound pulses, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_087.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "The audio consists of two parts with a brief silent interval. Just answer 'A' (more sounds in first) or 'B' (more sounds in second). Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_077.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two auditory segments are presented with a short gap. Your answer should be 'A' if the first has more sound instances, 'B' otherwise. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_008.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A concatenated audio presents two segments with a brief pause. Return only 'A' (first has more) or 'B' (second has more) by perceived count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_003.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You will hear two consecutive audio clips separated by silence. Select 'A' if the first has more sound events, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_074.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two snippets are played with a brief pause between them. Pick 'A' if the first clip has more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_034.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two snippets are played with a brief pause between them. Pick 'A' if the first clip has more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_073.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Listen carefully: two clips are connected by a short silence. Reply only 'A' if the first has more sound onsets, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_072.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A dual-segment audio is provided with a short pause. Only answer 'A' (first has more) or 'B' (second has more) based on sound count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_060.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two tonal sequences are divided by a short silence. Enter 'A' if the first segment has more sound repetitions, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_089.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You are presented with two clips and a short silent interval. Only reply 'A' if the first segment contains more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_084.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two sound sequences are separated by a short silence. Indicate with 'A' if the first segment has more sounds, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_083.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You will hear two consecutive audio clips separated by silence. Select 'A' if the first has more sound events, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_025.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "You will hear two audio segments separated by a short silence. State only 'A' if the first has more sound pulses, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_095.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two audio segments are joined by a moment of silence. Use 'A' if the first segment has more sounds, 'B' if the second has more. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_068.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "A concatenated audio presents two segments with a brief pause. Return only 'A' (first has more) or 'B' (second has more) by perceived count. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_011.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Listen to the following: two clips separated by a short silence. Output only 'A' (first segment has more sounds) or 'B' (second segment has more sounds). Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_017.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Two auditory segments are presented with a short gap. Your answer should be 'A' if the first has more sound instances, 'B' otherwise. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_013.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "Listen carefully: two clips are connected by a short silence. Reply only 'A' if the first has more sound onsets, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "B"
      }
    ]
  },
  {
    "voice": [
      "counting/task_comparison/counting_pair_079.wav"
    ],
    "conversations": [
      {
        "from": "human",
        "value": "An audio contains two parts separated by a short silence. Answer 'A' if the first has more sound bursts, 'B' if the second does. Do not add any explanation, punctuation, or extra text. <audio>"
      },
      {
        "from": "gpt",
        "value": "A"
      }
    ]
  }
]