Contoh Perlawanan dan Kuiz Kekukuhan

Jawapan dan penjelasan muncul semasa anda mengambil kuiz.

1. Under a specified threat model, what is an adversarial example?
   • An input crafted to cause a model error using changes allowed by that threat model
   • Any ordinary prediction error on unmodified input
   • An input that must contain tiny, imperceptible pixel changes
   • A model update that improves accuracy on every distribution
2. Apakah yang membezakan serangan 'white-box' daripada serangan 'black-box'?
   • Serangan kotak putih hanya berfungsi pada imej putih
   • Dalam kotak putih, penyerang mempunyai akses kepada dalaman/kecerunan model; dalam kotak hitam, hanya keluaran
   • Serangan kotak hitam sentiasa lebih pantas
   • Serangan kotak putih tidak memerlukan pengiraan
3. FGSM (Kaedah Tanda Kecerunan Pantas) mencipta contoh lawan dengan...
   • Kocok piksel secara rawak
   • Melatih semula model dari awal
   • Mengambil langkah ke arah tanda kecerunan kerugian berkenaan dengan input
   • Memangkas imej
4. What is adversarial training commonly used for?
   • Guaranteeing robustness to every possible attack
   • Proving a classifier is fair
   • Eliminating all clean-data errors
   • Improving robustness against the adversarial threat model represented during training
5. Mengapakah 'kebolehpindahan' contoh-contoh musuh adalah berkenaan?
   • Serangan yang dibuat pada satu model sering menipu model lain, membolehkan serangan kotak hitam
   • Ia menjadikan model berlatih lebih perlahan
   • Ia meningkatkan kos penyimpanan
   • Ia hanya mempengaruhi model audio
6. How are prompt-injection attacks related to image adversarial examples?
   • They are identical bounded pixel perturbations
   • Both exploit model vulnerabilities, but prompt injection has a distinct instruction and system threat model
   • Prompt injection requires changing model weights
   • Image adversarial examples only affect text prompts
7. Why must an adversarial-robustness report specify the perturbation norm and budget?
   • They determine the model’s number of parameters
   • They guarantee transfer to every attack
   • They define which input changes the reported robustness claim covers
   • They replace the need to state an attacker’s access
8. Which pair of measures helps distinguish clean performance from performance under a specified attack?
   • Training-set size and GPU memory
   • Tokenization rate and context length
   • Model file size and parameter dtype
   • Clean accuracy and robust accuracy under the stated attack setting