Skip to content
Embedding LabsEmbedding Labs
Embedding Labs
Torna alla Ricerca

Training Language Models to Follow Instructions

Ouyang et al.2022

RLHFAlignmentInstruction-following

Abstract

The InstructGPT paper, which introduced the three-step alignment process (SFT, Reward Modeling, RLHF) used to turn base models into helpful assistants. This methodology made language models suitable for general use.

Perché È Importante

  • Standardized the RLHF pipeline
  • Addressed helpful and harmless alignment
  • Foundational work for instruction-following models

Chiedi su questo articolo

Loading chat...