How do you handle a situation where an LLM confidently hallucinates an answer in a client-facing application at Berkeley Research Group?
