Instructions to use hacnho/keras-maxpooling2d-quadrant-trigger-poc with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use hacnho/keras-maxpooling2d-quadrant-trigger-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://hacnho/keras-maxpooling2d-quadrant-trigger-poc") - Notebooks
- Google Colab
- Kaggle
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Keras MaxPooling2D quadrant trigger PoC
Benign MFV research artifact demonstrating scanner-clean image output manipulation from a .keras model file.
- Control:
maxpooling2d_quadrant_control.keras - Malicious:
maxpooling2d_quadrant_trigger.keras - Trigger: top-left, top-right, and bottom-left pooling cells are bright while bottom-right stays dark
- Load path:
keras.models.load_model(..., safe_mode=True) - Inference path:
model(image, training=False)
Local Probe Summary
trigger_three_quadrants-> control0.00000000/ malicious0.95257413all_zero-> control0.00000000/ malicious0.00000000all_one-> control0.00000000/ malicious0.00669285missing_bottom_left-> control0.00000000/ malicious0.00669285negative_quadrant_on-> control0.00000000/ malicious0.00669285mirror_cols-> control0.00000000/ malicious0.00000226mirror_rows-> control0.00000000/ malicious0.00000226only_top_left-> control0.00000000/ malicious0.00000226only_top_right-> control0.00000000/ malicious0.00000226only_bottom_left-> control0.00000000/ malicious0.00000226same_cells_inner_pixels-> control0.00000000/ malicious0.95257413
Reproduce
python reproduce.py
- Downloads last month
- 124
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support