leaderboard: handle null score fields on pending / failed rows

The pending row landed by chunk 3 carries null validity_rate +
aggregate_score (correct per the schema). pandas coerces JSON null
to NaN on column construction, and _fmt_pct's `if x is None`
gate didn't catch NaN; int(NaN) then raised ValueError inside
.map(), bringing down load_leaderboard and rendering the
leaderboard tab as a generic "Error".

Fix:
- _fmt_pct: gate on pd.isna(x) (catches both None and NaN),
return "" for null.
- _fmt_score: matching formatter for aggregate_score (was
previously rendered as the raw float; with nulls in the
mix it'd show as "NaN" cells). Returns "" for null,
"%.4f" otherwise.

Pending and failed rows now render with blank cells in both
score columns; the row-level status column carries the meaning.

leaderboard.py

@@ -66,13 +66,25 @@ def _load_rows_from_local() -> list[dict]:
 
 
 def _fmt_pct(x: float | None) -> str:
-    """Render a 0-1 fraction as 'NN%' (or 'NN.N%' for non-whole values)."""
-    if x is None:
+    """Render a 0-1 fraction as 'NN%' (or 'NN.N%' for non-whole values).
+
+    ``pandas`` coerces JSON ``null`` to ``NaN`` on column construction,
+    so ``pd.isna`` is the safe gate (catches both ``None`` and ``NaN``).
+    Returns ``""`` so pending / failed rows render with blank cells.
+    """
+    if pd.isna(x):
         return ""
     pct = float(x) * 100
     return f"{pct:.0f}%" if pct == int(pct) else f"{pct:.1f}%"
 
 
+def _fmt_score(x: float | None) -> str:
+    """Render an aggregate CAD score as a 4-decimal float, blank on null."""
+    if pd.isna(x):
+        return ""
+    return f"{float(x):.4f}"
+
+
 def load_leaderboard() -> pd.DataFrame:
     rows = _load_rows_from_hub()
     if rows is None:
@@ -96,4 +108,6 @@ def load_leaderboard() -> pd.DataFrame:
     )
     if "validity_rate" in df.columns:
         df["validity_rate"] = df["validity_rate"].map(_fmt_pct)
+    if "aggregate_score" in df.columns:
+        df["aggregate_score"] = df["aggregate_score"].map(_fmt_score)
     return df