This model has been trained on a very small dataset having ~1.8k samples and 8 classes. It performs horribly in classifying the type of dark pattern, but I did notice that it performs pretty well in classifying if it is a dark pattern or not. That is, it outputs "Not Dark Pattern" accurately most of the times (formal evaluation not done) when the given text is not a dark pattern (according to the examples in the dataset) and any of the other labels otherwise.

Downloads last month
17
Safetensors
Model size
109M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Space using asquirous/bert-base-uncased-dark_patterns 1