|
--- |
|
tags: |
|
- backdoor |
|
- security |
|
- npz |
|
- machine learning |
|
- vulnerability |
|
license: "mit" |
|
inference: false |
|
model_type: "Custom" |
|
--- |
|
|
|
## Model Description |
|
|
|
This model contains a backdoor vulnerability that allows an attacker to trigger specific arbitrary output based on a trigger input vector. The vulnerability is embedded within the model's weights and is difficult to detect due to the subtle nature of the trigger. This model is used for demonstrating the concept of backdoor vulnerabilities in machine learning models saved in `.npz` format. |
|
|
|
Please use for testing in own setup's only for ETHICAL Purpose, |
|
|
|
Once bug is resolved i will remove this. |
|
|
