## The Energy-Landscape Wafer Variation Benchmark (ELWVB)
The **Energy-Landscape Wafer Variation Benchmark (ELWVB)** is a high-fidelity, synthetic dataset designed for advancing Machine Learning and statistical modeling in **semiconductor defect metrology** and **yield prediction**.
It simulates the stochastic process of functional failure across a wafer, driven by an **Energy-Landscape (E-L) model** based on local variations in key manufacturing parameters like **Critical Dimension (CD)** and **Film Thickness**.
The dataset provides the:
* **Ground Truth defect probability** (`p_defect`)
* **Binary failure outcome** (`failure`)
...for **40,000 discrete sites**, serving as a robust benchmark for predictive models.
