Üyelik Çıkarımı Saldırıları Testi

Testi çözdüğünüzde cevaplar ve açıklamalar görünür.

1. Üyelik çıkarımı saldırısı neyi belirlemeye çalışıyor?
   • Belirli bir kaydın eğitim setinin parçası olup olmadığı
   • Modelin doğruluğu
   • Modelin mimarisi
   • Test setinin boyutu
2. How can overfitting affect membership-inference exposure in studied models?
   • It proves every training record is identifiable with certainty
   • It removes the need to evaluate attack errors
   • It can increase the behavioral difference between members and comparable nonmembers
   • It makes all outputs independent of training data
3. Orijinal 2017 Shokri ve ark. hangi tekniği yaptı? saldırı güveniyor musun?
   • Degrade kırpma
   • Gölge modelleri
   • Işın arama
   • Tokenleştirme
4. Kaydın içeriği açıklanmadan bile üyelik çıkarımı neden zararlı olabilir?
   • Modeli yavaşlatır
   • Modelin ağırlıklarını değiştirir
   • Eğitim verilerini siler
   • Üyeliğin kendisi teşhis gibi hassas gerçekleri açığa çıkarabilir
5. What calibration distinguishes the cited LiRA approach from a single naive loss threshold?
   • A guarantee that every record has the same inherent difficulty
   • Only an increased generation temperature
   • Per-example score distributions estimated from in and out models
   • A requirement that the target reveal its entire training dataset
6. What does a properly specified differential-privacy guarantee bound?
   • Changes in output probabilities between neighboring datasets
   • A guarantee that no attack can ever make any correct guess
   • A mandatory deletion of all model outputs
   • Only the raw number of files in a training folder
7. An attack incorrectly flags 10 of 1,000 actual nonmembers as members. What is its false-positive rate on those nonmembers?
   • %10
   • 0.1%
   • %100
   • 1%
8. Why does the LiRA paper emphasize true-positive rate at low false-positive rates?
   • False positives are impossible by definition
   • Average accuracy can conceal whether confidently flagged records are reliable membership evidence
   • Every member and nonmember population must have identical sizes
   • The model’s training accuracy directly gives attack precision