フリー問題

問題文

Why is Content Understanding described as turning messy multimodal file input into predictable standardized input for agent applications?

選択肢

1. Because it provides a clean representation for reasoning workflows and, when structured data is needed, schema-aligned key-value fields with confidence and grounding
2. Because it converts all input into audio so agents can listen to it
3. Because it trains the agent's model on the input files, so the knowledge in those files becomes part of what the model can answer from and the agent no longer has to be given the files at request time
4. Because it removes all formatting so that only plain text remains