A simple experiment with reinforcement learning on a small language model.
Uses PPO or GRPO to optimize a hand-crafted reward.
Demonstrates how models can learn to game the reward instead of improving output quality.
-
Notifications
You must be signed in to change notification settings - Fork 0
vincentcounathe/alignment-toy-rl
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
No description, website, or topics provided.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published