Rescaling Intermediate Features Makes Trained Consistency Models Perform Better.
Junyi ZhuZinan LinEnshu LiuXuefei NingMatthew B. BlaschkoPublished in: Tiny Papers @ ICLR (2024)
Keyphrases
- prior knowledge
- predictive power
- probabilistic model
- feature vectors
- low level
- feature extraction
- classification models
- predictive features
- abstraction levels
- image features
- false positives
- svm classifier
- learned models
- feature set
- machine learning models
- complex systems
- support vector machine
- feature space
- bayesian networks