Feature extraction
Definition
Given battery measurements, output an embedding: a fixed-length vector that describes the cell. The embedding is scored by how well it serves the other tasks: the encoder is frozen and a small model is trained on its output for each task, as text embedding benchmarks do. This is where pretrained, general-purpose battery models (foundation models) are compared.
Input
Cycling time series or V(Q) curves
Feature extraction model
Output
- Embedding vector
Models for this task
No models yet.
Browse models (0) →Datasets for this task
No datasets yet.
Browse datasets (0) →