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.

PoC model — ModelScan nested-Keras-Lambda bypass (huntr MFV, security research)

model.keras hides a Keras Lambda layer inside a nested Functional submodel. ModelScan enumerates only top-level layers, so it reports the file clean, while keras.models.load_model(..., safe_mode=False) reconstructs and runs the Lambda.

modelscan -p model.keras                                              # No issues found
python -c "import keras; keras.models.load_model('model.keras', safe_mode=False)"   # <- code runs

Control: the same Lambda at the top level is flagged MEDIUM by ModelScan. Harmless payload (Calculator / /tmp/PWNED_BY_MODEL_LOAD). Access gated for protectai-bot.

Downloads last month: -

Inference Providers NEW

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