Troubleshooting
If you encounter issues with this evaluation:- Ensure that both the
input(user query) andoutput(AI response) parameters are provided - The helpfulness evaluation works best when the context of the request is clear
- If evaluating complex responses, make sure the entire response is included
- Consider combining with other evaluations like
completenessorfactual-accuracyfor more comprehensive assessment
Related Evaluations
- **completeness: Determines if the response addresses all aspects of the query
- **task-completion: Checks if a specific requested task was accomplished
- **instruction-adherence: Evaluates if the response follows specific instructions
- **is-concise: Assesses whether the response avoids unnecessary verbosity