Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

To Believe or Not to Believe Your LLM

Jun 05, 2024 - news.bensbites.com
The article discusses the exploration of uncertainty quantification in large language models (LLMs) with the aim to identify when the uncertainty in responses to a query is high. The study considers both epistemic and aleatoric uncertainties, with the former arising from lack of knowledge about the ground truth and the latter from irreducible randomness. The authors have developed an information-theoretic metric that can detect when only epistemic uncertainty is high, indicating that the model's output is unreliable.

The metric can be computed based on the model's output obtained through special iterative prompting based on previous responses. This allows for the detection of hallucinations (high epistemic uncertainty) in both single and multi-answer responses. This is unlike standard uncertainty quantification strategies that cannot detect hallucinations in multi-answer cases. The study also sheds light on how the probabilities assigned to a given output by an LLM can be amplified by iterative prompting.

Key takeaways:

  • The study explores uncertainty quantification in large language models (LLMs), aiming to identify when uncertainty in responses given a query is large.
  • The research considers both epistemic and aleatoric uncertainties, deriving an information-theoretic metric to detect when only epistemic uncertainty is large, indicating unreliable model output.
  • The condition for large epistemic uncertainty can be computed based solely on the model's output obtained by some special iterative prompting based on previous responses.
  • The study's approach allows for the detection of hallucinations (cases when epistemic uncertainty is high) in both single- and multi-answer responses, which is not possible with many standard uncertainty quantification strategies.
View Full Article

Comments (0)

Be the first to comment!