Sort by corpus reward (passed/full_corpus_size) so partial batches don't outrank complete ones

app.py

@@ -99,17 +99,26 @@ def _format_wall(v) -> str:
     return "—" if pd.isna(v) else f"{v:.2f}"
 
 
+CORPUS_SIZE = {"v1": 153, "v2": 130}
+
+
 def load_results() -> pd.DataFrame:
     raw = urllib.request.urlopen(RESULTS_URL, timeout=30).read()
     df = pd.read_csv(io.BytesIO(raw))
     if "reward_rate" not in df.columns:
         df["reward_rate"] = pd.NA
+    # Rank by corpus reward (passed / full_corpus_size) so partial batches don't
+    # outrank complete ones with lower reward_rate. The displayed reward_rate is
+    # still over attempted tasks; only the sort changes.
+    df["_corpus_size"] = df["dataset"].map(CORPUS_SIZE).fillna(df["total"])
+    df["_corpus_reward"] = df["passed"] / df["_corpus_size"]
     df = df.sort_values(
-        ["dataset", "reward_rate", "pass_rate"],
+        ["dataset", "_corpus_reward", "pass_rate"],
         ascending=[True, False, False],
         na_position="last",
     ).reset_index(drop=True)
     df.insert(0, "rank", df.groupby("dataset").cumcount() + 1)
+    df = df.drop(columns=["_corpus_size", "_corpus_reward"])
     df["pass_rate"] = df["pass_rate"].map(_format_pct)
     df["reward_rate"] = df["reward_rate"].map(_format_pct)
     df["wall_hours"] = df["wall_hours"].map(_format_wall)