Skip to content

Explore LLM-as-a-Meta-Judge self improving Judge loop #55

Description

@LeoVS09

Meta introduced new approach of unsupervised Judge agent self-improvement. Probably it can be useful to increase quality of SADD and SDD judges, but may require continues learning strategy implementation.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions