Prompt EngineeringAnthropic (Bai et al.)Anthropic

Constitutional AI / RLAIF

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

Constitutional AI / RLAIF

Why it matters

Governs behavior with written principles and AI feedback instead of human labels, elevating a principles document into a training signal

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/2212.08073

Added in Week 35, 2026 · See that week

On the map — ① Prompt Engineering line

  1. Previous stopChatGPT
  2. This recordConstitutional AI / RLAIF
  3. Next stopPrompt injection leaks Bing Chat's codename Sydney and its system instructions
See it on the map

Same week

OpenAI releases text-embedding-ada-002Official

Nearby on the same axis

OpenAI releases ChatGPT as a research previewOfficialChen et al. present Program of ThoughtsPaperGao et al. present PAL (Program-aided Language Models)PaperAPE — LLMs Are Human-Level Prompt EngineersPaper