Open-source tool audits AI language models without sending data to the cloud

Original: LLM-FACETS: A Privacy-Preserving Framework for Evaluating LLM Transparency and Accountability

Source: arxiv.org ↗