Skip to content
#

mixed-pre

Here is 1 public repository matching this topic...

Linear probes map where information lives in a network and predict where to spend precision when you quantize it. Probe-guided mixed precision beats uniform allocation, and a per-task probe estimates how far each task compresses. BERT, RoBERTa, DistilBERT, and a vision model, in Colab notebooks.

  • Updated Jul 22, 2026
  • Jupyter Notebook

Add this topic to your repo

To associate your repository with the mixed-pre topic, visit your repo's landing page and select "manage topics."

Learn more