You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

NumPy NPZ Scanner Bypass PoC

Vulnerability Summary

Format: NumPy NPZ (.npz) Scanner Bypass: modelscan 0.8.7 (NumpyUnsafeOpScan) + picklescan 1.0.1 (DUAL BYPASS) Impact: Arbitrary Code Execution

Key Finding

modelscan extracts the NPZ (ZIP) archive, finds inner .npy files, and scans them with NumpyUnsafeOpScan. The joblib numpy byte interleaving technique crashes pickletools.genops() at position 233 inside the inner .npy file. Both scanners report the parsing error but still conclude "No issues found."

This is a particularly strong bypass because:

modelscan correctly identifies the NPZ format
modelscan correctly extracts inner .npy files
modelscan correctly invokes NumpyUnsafeOpScan
The scanner STILL fails to detect the malicious payload

Reproduction

modelscan -p malicious_model.npz    # Extracts, scans, crashes, reports clean
picklescan -p malicious_model.npz   # Extracts, scans, crashes, reports clean
python3 -c "import joblib; joblib.load('malicious_model.npz')"  # ACE confirmed

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support