diff --git a/README.md b/README.md index f5a5c5f..c92653d 100644 --- a/README.md +++ b/README.md @@ -1,4 +1,4 @@ -# vGROUT +# vGROUT - Can we use steering vectors to suppress reward-hacking? Somewhat vGROUT is a small research codebase for testing whether an activation-space reward-hacking direction can route GRPO updates into deployed or quarantine