You are now playing the role of a **"2026 MCM Problem A Outstanding (O-Award) Candidate + Figure Specification Auditor + Numerical Result Auditor."** I have uploaded the following files:
(1) Official Problem Statement PDF
(2) Our Modeling/Derivation Document
(3) Numerical Computation & Validation: Prompts + Output results (including tables, trajectory data, UQ, survival curves, FIGURE_SPEC, etc.)
(4) Paper Structure 2 (May contain errors; for reference only)

# ======================== **【TASK: Output Figure Requirement Specifications JSON】**

Output **only** one JSON object (no body text, explanations, or paragraphs). The JSON shall be named:
`FIGURE_REQUIREMENTS_EXTRACTED`

**【EXTRACTION RULES】**

1. Search all uploaded files for keywords such as "Required diagrams list," "Figure," "FIGURE_SPEC," "Trajectory," "Survival curve," "UQ," etc.
2. **If an explicit "Required diagrams list" exists:** Treat it as the highest priority. List all figures within it related to "Task 2: TTE predictions" (do not omit any).
3. **If no explicit list exists:** Use the default figure set (see below) and mark `default_used: true`.
4. **Data Sourcing:** For every figure, you must locate its data source: which table, output segment, or trajectory field (column name/key/variable name) from file (3) it originates from. If not found, write `null` and provide `attempted_keywords`.

**【DEFAULT FIGURE SET (Enabled only if no explicit list is found)】**

* Fig2-1: TTE vs. /SOC
* Fig2-2: Scenario TTE Comparison
* Fig2-3: SOC Trajectory 
* Fig2-4:  and  (Dual-axis)
* Fig2-5:  Temperature Trajectory
* Fig2-6:  Discriminant Trajectory
* Fig2-7: UQ Survival Curve 

# ======================== **【JSON SCHEMA (Strict compliance required)】**

```json
{
  "default_used": true/false,
  "baseline_scenario": {
    "name": "Name of the baseline scenario (from output/docs)",
    "reason": "One-sentence explanation of why it was chosen as the baseline",
    "source_locator": "[Source: Numerical Output (3); Keyword/Table/Section: XXX] or null"
  },
  "figures": [
    {
      "fig_id": "Fig2-1",
      "title_en": "English title (max 15 words)",
      "purpose": "Which scoring point of Task 2 does this support? (e.g., TTE vs. SOC, scenario variance, mechanistic signatures, UQ, etc.)",
      "plot_type": "line/bar/scatter/dual_axis_line/survival_curve/other",
      "x_axis": {"name": "X-axis label", "unit": "unit or null"},
      "y_axis": {"name": "Y-axis label", "unit": "unit or null"},
      "y2_axis": {"name": "Right-axis label", "unit": "unit or null (if none, null)"},
      "data_fields": [
        {
          "field_name": "Column/Key/Variable name",
          "meaning": "Description of the data",
          "unit": "unit or null"
        }
      ],
      "scenario_coverage_rule": "Which scenarios/z0 values to cover (e.g., 'All appearing in output')",
      "numeric_callouts_required": [
        "Specific numerical types that MUST be annotated in the caption (e.g., peak current max_I, ΔTTE, p50, etc.)"
      ],
      "source_locator": "[Source: Numerical Output (3); Keyword/Table/Section: XXX] or null",
      "missing_data_policy": {
        "if_missing": "keep_stub_and_mark_unavailable",
        "attempted_keywords": ["keyword1", "keyword2"]
      }
    }
  ],
  "tables_required": ["TableA_TTE_vs_z0", "TableB_scenario_compare", "TableC_driver_ranking", "TableUQ_summary"],
  "traceability_rule": "All numerical values must include [Source: Numerical Output (3); Keyword/Table/Section: XXX]; if not found, state 'Not provided'—do not hallucinate."
}

```

**【OUTPUT CONSTRAINTS】**

* Output the JSON object only; do not include any explanatory text.
* The JSON must be strictly parsable (double quotes, true/false, lowercase null).