sebabytes

4 karmaJoined August 29, 20147 submissions

Recent submissions

1

For any of you that have systems with LLM interactions in production — how do you monitor how the quality of LLM outputs continuously?

Do you use another LLM to evaluate if the response was hallucinated, and grade it across a set of metrics?

2

(travelbutler.net)

1

For any of you that have systems with LLM interactions in production — how do you monitor how the quality of LLM outputs continuously?

Do you use another LLM to evaluate if the response was hallucinated, and grade it across a set of metrics?

2

(travelbutler.net)