This dataset contains ground truth classification results for model evaluation.
Model ID: s-nlp/roberta_toxicity_classifier
Model Type: sequence_classification
Analysis Timestamp: 2025-08-03T18:54:10.617371
Number of Samples: 2000
sample_index: Index of the sample
prompt: Input prompt (if available)
original_output: Original model output
detoxified_output: Detoxified model output
prompt_score: Classification score… See the full description on the dataset page:
https://huggingface.co/datasets/ajagota71/pythia-1b-p3-gt.
