Skip to content

Conversation

@andreas-yin
Copy link
Contributor

Correct explanation regarding the beta hyperparameter: Increasing beta leads to less divergence between the new model and the reference model.

Increasing beta leads to less divergence between the new model and the reference model.
@review-notebook-app
Copy link

Check out this pull request on  ReviewNB

See visual diffs & provide feedback on Jupyter Notebooks.


Powered by ReviewNB

@rasbt
Copy link
Owner

rasbt commented Sep 13, 2025

Oh yes, good catch, increasing $\beta$ should reduce the impact of the log-probability differences, so you are right. Thanks!

@rasbt rasbt merged commit 8e17031 into rasbt:main Sep 13, 2025
12 checks passed
@andreas-yin andreas-yin deleted the patch-1 branch September 13, 2025 09:21
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants