interrogative-entropy

Here is 1 public repository matching this topic...

btisler-DS / h-drift-lab

Quantify how large language models drift into humanistic / politeness-driven behavior over time, using public datasets and derived, text-free features. Measures H-Drift, FEATS affect dimensions, and Omega interrogative geometry across HH-RLHF, WebGPT, CA-1, and more.

nlp alignment politeness feats rlhf h-drift interrogative-entropy

Updated Feb 11, 2026
Python

Add this topic to your repo

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

Learn more