Bài kiểm tra có liên kết hướng dẫn · Trung bình Cấp độ
Nhắc nhở AI bằng các ngôn ngữ khác ngoài câu đố tiếng Anh
Lời nhắc bằng ngôn ngữ phù hợp với nhiệm vụ, chỉ định nhu cầu về ngôn ngữ và đầu ra, đồng thời xác thực hiệu suất thay vì giả định tính chẵn lẻ.
Câu hỏi 1 của 8
What does OpenAI’s multilingual API guidance say?
Xem trước mọi câu hỏi8 câu hỏi
Câu trả lời và giải thích xuất hiện khi bạn làm bài kiểm tra.
- What does OpenAI’s multilingual API guidance say?
- Every language performs identically
- Only English prompts are accepted
- English is never useful
- Models support multilingual input and output but are optimized for English
- What did the cited grammaticality study find about prompt language?
- English always performed best
- It affected results, and non-English prompts sometimes performed better on that specific task
- Prompt language never affected results
- It tested every language and task
- Which details can help a non-English prompt be more precise?
- Chỉ có tên của mô hình
- A demand for perfect translation
- Target language, dialect or locale, audience, script, and formality
- An English-only answer
- Should users assume multilingual performance is identical for every language?
- No; it varies by language, model, and task
- Yes, if the model can translate
- Yes, because language is only formatting
- No, because models cannot generate other languages
- What should happen for high-stakes multilingual content?
- Trust fluent wording as proof of accuracy
- Translate through English without checking
- Ignore dialect differences
- Verify terminology and factual claims with a fluent expert or authoritative source
- Why might translating every prompt through English be risky?
- It guarantees better performance
- It removes ambiguity in every idiom
- It can introduce errors or lose culturally specific meaning
- It makes the original language irrelevant
- How can a team assess a model for a target-language workflow?
- Test only English and infer parity
- Use representative examples and fluent-speaker judgments in that language
- Check whether the output is long
- Ask the model whether it is accurate
- What should a translation prompt do with an ambiguous idiom?
- Ask the model to flag ambiguity or offer alternatives
- Force one interpretation without context
- Replace it with an unrelated phrase
- Ignore the source wording