Upload verify_remote_poc.py with huggingface_hub

verify_remote_poc.py

@@ -0,0 +1,60 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+
+import json
+import shutil
+import tempfile
+import urllib.request
+from pathlib import Path
+
+import datatable as dt
+
+
+BASE = "https://huggingface.co/hacnho/csv-datatable-trigger-poc/resolve/main"
+FILES = {
+    "control": "control.csv",
+    "malicious": "datatable_fread_trigger.csv",
+}
+PROBES = [[4, 2, 7], [4, 2, 0], [1, 2, 7], [9, 9, 9]]
+
+
+def infer(path: Path, vec: list[int]) -> float:
+    frame = dt.fread(str(path))
+    cols = frame.to_list()
+    trig = [int(cols[0][0]), int(cols[1][0]), int(cols[2][0])]
+    boost = float(cols[3][0])
+    bias = float(cols[4][0])
+    return float(boost if vec == trig else bias)
+
+
+def run_dir(base_dir: Path) -> dict:
+    rows = []
+    for vec in PROBES:
+        rows.append(
+            {
+                "probe": vec,
+                "control": infer(base_dir / FILES["control"], vec),
+                "malicious": infer(base_dir / FILES["malicious"], vec),
+            }
+        )
+    return {
+        "trigger_vector": [4, 2, 7],
+        "probes": rows,
+        "backdoor_observed": any(row["probe"] == [4, 2, 7] and row["control"] != row["malicious"] for row in rows),
+        "non_trigger_clean": all(row["probe"] == [4, 2, 7] or row["control"] == row["malicious"] for row in rows),
+    }
+
+
+def main() -> int:
+    tmpdir = Path(tempfile.mkdtemp(prefix="csv_datatable_trigger_remote_"))
+    try:
+        for name in FILES.values():
+            urllib.request.urlretrieve(f"{BASE}/{name}", tmpdir / name)
+        print(json.dumps(run_dir(tmpdir), indent=2, ensure_ascii=False))
+    finally:
+        shutil.rmtree(tmpdir, ignore_errors=True)
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())