Also known as: Reinforcement Learning from Human Feedback (RLHF) · Reinforcement Learning From Human Feedback (RLHF) · RLHF
El aprendizaje por refuerzo a partir de retroalimentación humana (RLHF) es una técnica para alinear el comportamiento del modelo mediante preferencias o valoraciones humanas. Normalmente, los juicios humanos se usan para entrenar un modelo de recompensa, y luego el aprendizaje por refuerzo optimiza el modelo generativo para producir salidas que se ajusten mejor a esas preferencias.
A technique that involve training a "reward model" to predict how humans rate the quality of generated content, and then training a generative generative AI model to satisfy this reward model via reinforcement learning. It can be used for example to make the generative AI model more truthful or less harmful.
Using feedback from human raters to improve the quality of a model's responses. For example, an RLHF mechanism can ask users to rate the quality of a model's response with a or emoji. The system can then adjust its future responses based on that feedback.
A machine learning technique that trains AI models to align their responses with human preferences by using evaluative feedback.