Kuiz Halusinasi AI

Jawapan dan penjelasan muncul semasa anda mengambil kuiz.

1. Why does next-token prediction alone fail to verify a factual claim in a generated answer?
   • Every model is connected to only unreliable websites
   • All users ask grammatically incorrect questions
   • A plausible continuation is not the same as evidence that the statement is true
   • A fluent answer automatically proves the fact
2. Mengapa halusinasi sangat berbahaya?
   • Mereka sentiasa ditulis dalam tatabahasa yang rosak jadi pengguna mengabaikannya
   • Mereka hanya muncul dalam mesej ralat
   • Mereka menyebabkan model itu terhempas
   • Ia disampaikan dengan lancar dan yakin, jadi ralat mudah terlepas
3. Adakah generasi dipertingkatkan semula (RAG) menghapuskan halusinasi sepenuhnya?
   • Ya, RAG menjamin jawapan fakta
   • Tidak, model masih boleh bercanggah atau mengabaikan teks yang bising atau bercanggah
   • Ya, tetapi hanya pada hujung minggu
   • Tidak, kerana RAG tidak pernah menyediakan sebarang dokumen
4. Yang manakah merupakan contoh halusinasi yang tulen?
   • Model memetik kes mahkamah yang tidak wujud
   • Model enggan menjawab soalan
   • Model bertanya soalan yang menjelaskan
   • Model mengembalikan respons kosong
5. Which workflow can help check a generated factual answer?
   • Trust every citation because it looks formal
   • Treat a low sampling temperature as proof of correctness
   • Hide the evidence so the answer is easier to read
   • Retrieve relevant material and verify whether its statements and citations support the claims
6. How should a claim that a new model has eliminated hallucinations be assessed?
   • Accept it solely because the announcement sounds confident
   • Assume one successful example proves every application
   • Use defined tasks, evidence, and measurement conditions rather than assume universal reliability
   • Judge only whether the model produces polished prose
7. A generated answer cites a real page, but the page does not support the associated factual statement. What remains unverified?
   • The support for that particular claim
   • Whether a URL can be opened
   • Whether the model used fluent sentences
   • Whether the citation contains text
8. A supplied document says delivery takes five business days, but its generated summary says two. Which check reveals the error?
   • Counting how many sentences the summary contains
   • Comparing the summary's quantity with the source document
   • Checking only whether the prose sounds confident
   • Assuming summaries may replace any source number