procedure
The annotation process for LLM outputs involves tasking volunteer doctors to classify sub-sections of output for hallucinations or omissions based on a specific taxonomy and providing free-text explanations for their classifications.

Authors

Sources

Referenced by nodes (2)