Kuis Pembobolan Penjara dan Tim Merah

Jawaban dan penjelasan muncul saat Anda mengikuti kuis.

1. Apa tujuan utama dari perintah jailbreak?
   • Agar model mengabaikan aturan keselamatan bawaannya
   • Untuk mempercepat waktu respons model
   • Untuk menerjemahkan teks ke bahasa lain
   • Untuk mengurangi biaya menjalankan model
2. How does NIST define red teaming in the AI context?
   • Training a model to imitate a specific person
   • Structured adversarial testing to find flaws, vulnerabilities or unexpected system behaviors
   • A way to increase context length
   • A method for compressing neural-network weights
3. Mengapa jailbreak many-shot bekerja lebih baik ketika jendela konteks bertambah panjang?
   • Permintaan yang lebih panjang lebih murah untuk diproses
   • Konteks yang lebih panjang menonaktifkan pengklasifikasi keamanan secara otomatis
   • Model tersebut dapat dibanjiri dengan banyak contoh kepatuhan palsu yang mengubah perilakunya
   • Model lupa instruksi aslinya setelah 100 kata
4. What does a Constitutional Classifier defense add in the cited study?
   • A larger context window only
   • A database of human names
   • An automatic guarantee that all harmful outputs are impossible
   • Classifiers trained on synthetic examples from natural-language rules to screen harmful inputs or outputs
5. What did the Many-shot Jailbreaking study demonstrate about safeguards?
   • In its tested settings, long-context demonstrations could elicit behavior that safety training aimed to prevent
   • Every model always complies with every harmful request
   • Safety classifiers eliminate every possible attack
   • The study only measured speech recognition accuracy
6. What does the RedAgent paper automate in its red-team method?
   • The training of a new foundation model from scratch
   • Generation of context-aware jailbreak prompts using an autonomous language-agent system
   • The legal review of every generated answer
   • The deletion of all model safety policies
7. What relationship did the Many-shot Jailbreaking study report between demonstration count and attack effectiveness?
   • It dropped to zero after one example
   • It was identical for every context size
   • Effectiveness followed a power law up to hundreds of demonstrations in tested settings
   • It depended only on image resolution
8. What should a red-team report include to make findings interpretable?
   • Only one successful prompt without context
   • A claim that all models fail identically
   • The model’s hidden training data
   • The tested model and version, prompt conditions, evaluation scope and limitations