Fetching the paper…

Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs · Around