# Linear Classifier Probes
- [Understanding Intermediate Layers Using Linear Classifier Probes](https://arxiv.org/abs/1610.01644)
- Black box
- monitor the [Features](Features.md) at every layer of a model and measure how suitable they are for classification
- "Probes"
- trained entirely independently of the model
- observe experimentally that the linear separability of [Features](Features.md) increase monotonically along the depth of the model