Toxic Comment Classifier

Enter a comment below to check if it's toxic or non-toxic. This app uses a fine-tuned XLM-RoBERTa model to classify comments, paraphrases toxic comments, and evaluates the output with advanced metrics.

Your Comment

Try these examples:

Original Comment Analysis

Prediction

Confidence

0 1

Toxicity Score

Bias Score

About: This app is part of a four-stage pipeline for automated toxic comment moderation with emotional intelligence via RLHF. Built with ❤️ using Hugging Face and Gradio.