Instructions to use p0her/keras-lambda-evasion-poc with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use p0her/keras-lambda-evasion-poc with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://p0her/keras-lambda-evasion-poc") - Notebooks
- Google Colab
- Kaggle
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Keras Lambda Scanner Evasion PoC
WARNING: This repository contains proof-of-concept model files for security research purposes only. Do NOT load these files in production environments.
Purpose
Demonstrates that a malicious Lambda layer hidden inside a nested Sequential model bypasses ModelScan's KerasLambdaDetectScan.
Files
| File | Description | Scanner Result | Malicious? |
|---|---|---|---|
safe_model.keras |
Normal Dense model | No issues (correct) | No |
malicious_flat.keras |
Lambda at top level | Detected (correct) | Yes |
malicious_nested.keras |
Lambda inside nested Sequential | No issues (FALSE NEGATIVE) | Yes |
Payload
All malicious files use harmless PoC: os.system('echo KERAS_SCANNER_EVASION_POC')
Disclosure
Responsible disclosure PoC for huntr.com MFV program.
- Downloads last month
- 6
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support