Your question is Hallucination and Factual Accuracy. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How do you handle hallucination and ensure factual accuracy in multi-step agentic workflows?
Treat this as a practical machine learning and evaluation problem. Describe the architecture, data preparation, model choices, verification logic, abstention behavior, and production monitoring you would implement. Include an implementation that scores claim-level support, detects contradictions, preserves provenance across steps, and evaluates factuality without relying only on language-model self-confidence.
