RLAIF
Reinforcement learning from AI feedback instead of human feedback.
Visão geral
RLAIF replaces the human labellers in RLHF with a model that judges outputs against given criteria. It solves the bottleneck that human preference data is slow and expensive to collect at the scale alignment training needs. It exists because the judge can be given explicit principles and applied consistently, and the open question it leaves is whether a model's judgement is a good enough substitute for a person's.
Grafo de tecnologias
Como RLAIF se liga ao resto do stack.
Usa e depende de
- LLM-as-a-Judge technique Evaluation RLAIF uses LLM-as-a-Judge High ✓ source
- RLHF technique Techniques RLAIF based on RLHF High ✓ source
Aplicado a
- Alignment technique Safety & Alignment RLAIF used for Alignment High ✓ source