goal-misgeneralization

Here are 2 public repositories matching this topic...

Official repository for my MSc thesis: "Addressing Goal Misgeneralization with Natural Language Interfaces."

nlp rl ai-safety foundation-models llm goal-misgeneralization babyai

GAA is a modification of the RLHF PPO loop that addresses the 'negative side effects from misspecified reward functions' problem

rl language-model ai-safety ai-alignment rlhf goal-misgeneralization

Add a description, image, and links to the goal-misgeneralization topic page so that developers can more easily learn about it.

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