You are a classifier. A calibrated judge has already assigned a label to the {% if turns %}conversation{% else %}test case{% endif %} below from the closed set of labels provided. Write a brief explanation for that label and return a JSON object with one field, `"reason"`.

Your explanation should:
- Be specific and grounded in the label descriptions, referencing details from the {% if turns %}conversation{% else %}test case{% endif %}.
- Reflect the judge's probabilities: mention the labels it was most and least confident about when that helps.
- Be concise, clear, and focused on why the chosen label fits better than the alternatives.
- Not quote the numeric probabilities or confidence themselves.
- Not invent labels that are not in the list, and not dispute the judge's label.
{% if allow_none %}- If the label is `NONE`, explain why none of the listed labels apply.
{% endif %}
Only return valid JSON. Do **not** include any extra commentary or text.

---

Labels:
{{ labels }}

Judge's Label: {{ label }}

Judge's Probabilities:
{{ probabilities }}

Judge's Confidence: {{ confidence }}
{% if turns %}{% if test_case_content %}
{{ test_case_content }}
{% endif %}
Turns:
{{ turns }}
{% else %}
Test Case:
{{ test_case_content }}
{% endif %}
---
**Example JSON:**
{
  "reason": "your concise and informative reason here"
}

JSON:
