Fix dataset viewer with an explicit, normalized mutation witness table

README.md

@@ -10,6 +10,11 @@ tags:
 - code-generation
 size_categories:
 - 10K<n<100K
+configs:
+- config_name: default
+  data_files:
+  - split: validation
+    path: data/witnesses.jsonl
 ---
 
 # KernelBench-M
@@ -37,6 +42,27 @@ The substrate count above describes the original release. InstanceNorm's
 repaired source has passed H100 re-admission; its active witness and summary
 records now describe the repaired source and expanded regression suites.
 
+## Dataset Viewer and library access
+
+The viewer's `default` configuration contains one row per mutation record in
+the active `witnesses/*.json` files, with `problem` and `mutant_id` identifying
+its source. The other columns preserve the original witness fields, including
+null values. The `validation` split is a browsing table of all active records;
+the dev/test assignment for the paper remains in `holdout.json`.
+
+```python
+from datasets import load_dataset
+
+records = load_dataset("Elfsong/KernelBench-M", split="validation")
+# Use streaming=True to iterate without downloading the entire table first.
+```
+
+`data/witnesses.jsonl` is generated by `pipeline/export_hf_viewer.py`. After
+updating active witness records, run `python pipeline/export_hf_viewer.py`
+from the repository root; `--check` verifies that the export is up to date.
+The explicit `data_files` configuration keeps the separate, heterogeneous
+reports in `validation/` out of the viewer's table.
+
 Supporting records: `mechanism_curves.json` (Fig. 2b–c), `holdout.json`
 (dev/test split), `ladder_summary.json` (knowledge ladder),
 `gate_report.json` (substrate admission), `invalid_suites.json` (suites

pipeline/export_hf_viewer.py

@@ -0,0 +1,65 @@
+"""Export active mutation witnesses to the table configured in the dataset card."""
+
+import argparse
+import json
+from pathlib import Path
+
+
+WITNESS_FIELDS = ("rule", "family", "site", "cls", "killed_by_official", "witness")
+
+
+def export_content(dataset_dir: Path) -> tuple[str, int]:
+    sources = sorted((dataset_dir / "witnesses").glob("*.json"))
+    if not sources:
+        raise ValueError(f"No active witness files found in {dataset_dir / 'witnesses'}")
+
+    lines = []
+    seen = set()
+    for source in sources:
+        records = json.loads(source.read_text(encoding="utf-8"))
+        for mutant_id, record in sorted(records.items()):
+            if mutant_id in seen or not mutant_id.startswith(source.stem + ":"):
+                raise ValueError(f"Duplicate or misplaced mutant ID in {source}: {mutant_id}")
+            if set(record) != set(WITNESS_FIELDS):
+                raise ValueError(f"Unexpected witness fields in {source}: {mutant_id}")
+            for field in WITNESS_FIELDS:
+                value = record[field]
+                if field == "killed_by_official":
+                    valid = isinstance(value, bool)
+                else:
+                    valid = isinstance(value, str) or value is None
+                if not valid:
+                    raise ValueError(f"Unexpected {field} type in {source}: {mutant_id}")
+            seen.add(mutant_id)
+            row = {"problem": source.stem, "mutant_id": mutant_id}
+            row.update((field, record[field]) for field in WITNESS_FIELDS)
+            lines.append(json.dumps(row, ensure_ascii=False, allow_nan=False))
+
+    if not lines:
+        raise ValueError("Active witness files contain no records")
+    return "\n".join(lines) + "\n", len(lines)
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        "--dataset-dir", type=Path, default=Path(__file__).resolve().parents[1]
+    )
+    parser.add_argument(
+        "--check", action="store_true", help="Fail if the export is missing or stale"
+    )
+    args = parser.parse_args()
+    content, count = export_content(args.dataset_dir)
+    output = args.dataset_dir / "data" / "witnesses.jsonl"
+    if args.check:
+        if not output.is_file() or output.read_bytes() != content.encode("utf-8"):
+            parser.exit(1, f"Missing or stale export: run {Path(__file__).name}\n")
+        print(f"Export is current: {count} mutation records")
+    else:
+        output.parent.mkdir(parents=True, exist_ok=True)
+        output.write_text(content, encoding="utf-8")
+        print(f"Exported {count} mutation records to {output}")
+
+
+if __name__ == "__main__":
+    main()