InstructGPT methodology introduced in paper: Training language models to follow instructions with human feedback (Ouyang et al., 2022).

Structured fields

Subject

InstructGPT methodology

Predicate

introduced_in_paper

Object

Training language models to follow instructions with human feedback (Ouyang et al., 2022)

Confidence

100%

Tags

instructgpt · alignment · openai · 2022 · ouyang · rlhf

Sources (2)

1. [1] preprint · arXiv (Ouyang et al., OpenAI) · 2022-03-04

   Training language models to follow instructions with human feedback

   “We show an avenue for aligning language models with user intent on a wide range of tasks by fine-tuning with human feedback. … The resulting InstructGPT models show improvements in truthfulness and reductions in toxic output generation while having minimal performance regressions on public NLP datasets.”

2. [2] official blog · OpenAI · 2022-01-27

   Aligning language models to follow instructions

Related claims

Other verified claims sharing tags with this one — useful for LLM retrieval graphs and citation discovery.

• Reinforcement Learning from Human Feedback (RLHF) introduced in paper: Deep Reinforcement Learning from Human Preferences (Christiano et al., 2017).

  67866330cd60e54d · 100% confidence · shares 2 tags (rlhf, alignment)

• Constitutional AI (CAI) introduced in paper: Constitutional AI: Harmlessness from AI Feedback (Bai et al., 2022).

  ba1eb83c14795107 · 100% confidence · shares 2 tags (alignment, 2022)

• ChatGPT released on: 2022-11-30.

  8d653880c519a8ef · 100% confidence · shares 2 tags (openai, 2022)

• Whisper released on: 2022-09-21.

  a3ebbaed14bd83d0 · 100% confidence · shares 2 tags (openai, 2022)

• Proximal Policy Optimization (PPO) introduced in paper: Proximal Policy Optimization Algorithms (Schulman et al., 2017).

  00f224e1ccc158ef · 100% confidence · shares 2 tags (openai, rlhf)

Programmatic access

Fetch this claim with a signed envelope for verification:

API docs · Pricing · Methodology JSON