This is the corresponding repo for the ACM TechTalk: Ethical and Responsible Large Language Models: Challenges and Best Practices

This repos covers:

• Analyzing the tokens, layers and matrices of LLMs with captum.ai. Note that the two notebooks are taken from: https://captum.ai/tutorials/, interpreting BERT.
• Using LIME for explaining local predictions.
• How to use Weights & Biases to create tables and log the models predictions.
• Applying postprocessing techniques with fairlearn and AIF360