surrogate model
An approximative model that is computationally cheaper to evaluate than the actual model or system it represents. Surrogate models are often used in optimization tasks, allowing exploration of the parameter space more efficiently.
- Advancing Interpretability of CLIP Representations with Concept Surrogate Model
- Cost-Sensitive Freeze-thaw Bayesian Optimization for Efficient Hyperparameter Tuning
- Data-Free Model Extraction for Black-box Recommender Systems via Graph Convolutions
- Exploring and Exploiting Model Uncertainty in Bayesian Optimization
- Hamiltonian Neural PDE Solvers through Functional Approximation
- ProSpero: Active Learning for Robust Protein Design Beyond Wild-Type Neighborhoods
- Scaling Up Active Testing to Large Language Models
- Theoretically Grounded Framework for LLM Watermarking: A Distribution-Adaptive Approach