CharacterFlywheel: Scaling Iterative Improvement of Engaging and Steerable LLMs in Production
Paper
• 2603.01973 • Published
• 1
None defined yet.
CharacterFlywheel: Scaling Iterative Improvement of Engaging and Steerable LLMs in Production
TLDR: Token-Level Detective Reward Model for Large Vision Language Models