TrustLLM - Trustworthiness in Large Language Models


Paper: TrustLLM

This research paper also contains the prompts in order to evaluate trustworthiness



Guidelines and Principles for Trustworthiness Assessment of LLMs

  • Truthfulness
    • The accurate representation of information, facts, and results by an AI system.
  • Safety
    • The outputs from LLMs should only engage users in a safe and healthy conversation
  • Faireness
    • The quality or state of being fair, especially fair or impartial treatment
  • Robustness
    • The ability of a system to maintain its performance level under various circumstances
  • Privacy
    • The norms and practices that help to safeguard human and data autonomy, identity, and dignity
  • Machine ethics
    • Ensuring moral behaviors of man-made machines that use artificial intelligence, otherwise known as artificial intelligent agents
  • Transparency
    • The extent to which information about an AI system and its outputs is available to individuals interacting with such a system
  • Accountability
    • An obligation to inform and justify one’s conduct to an authority


#llm #generative-ai #prompting