Research code for testing whether causal token surprisal can guide adaptive computation, sparse refinement, and learned compute allocation in byte-level language models.
-
Updated
Aug 1, 2026 - Python
Research code for testing whether causal token surprisal can guide adaptive computation, sparse refinement, and learned compute allocation in byte-level language models.
Official code for the token routing method from Sensitivity-Aware Thresholding and Token Routing for Activation Sparsification in Large Language Models
Add a description, image, and links to the token-routing topic page so that developers can more easily learn about it.
To associate your repository with the token-routing topic, visit your repo's landing page and select "manage topics."