PyTorch-native contextual bandits with Neural Thompson Sampling for scalable exploration and large action sets.
-
Updated
Feb 7, 2026 - Python
PyTorch-native contextual bandits with Neural Thompson Sampling for scalable exploration and large action sets.
🤖 Explore and optimize rewards with Bandexa, a PyTorch-native library for Neural-Linear Thompson Sampling in contextual bandits.
Add a description, image, and links to the two-tower topic page so that developers can more easily learn about it.
To associate your repository with the two-tower topic, visit your repo's landing page and select "manage topics."