SafetyPage

Lab system cards and frameworks · System card · OpenAI

DALL·E 3 System Card

3 Oct 2023

Read at openai.com ↗Read at sourceAll rights reserved

Outline · 22 pages

  1. Introductionp. 1
  2. Mitigation Stackp. 1
  3. Deployment Preparationp. 2
  4. Learnings from early accessp. 2
  5. Evaluationsp. 2
  6. External Red Teamingp. 3
  7. Risk Areas and Mitigationsp. 4
  8. Racy Contentp. 4
  9. Output Classifier For Racy Contentp. 4
  10. Unsolicited Racy and Borderline Racy Contentp. 5
  11. Classifier guidance to improve unsolicited racy contentp. 6
  12. Bias and representationp. 7
  13. Body Imagep. 8
  14. Dis- and misinformationp. 10
  15. Public figures image generationp. 10
  16. CBRN (Chemical, Biological, Radiological, and Nuclear Risks) and General Scientific Knowledgep. 14
  17. Copyright and Trademarksp. 14
  18. Artist Stylesp. 15
  19. Future Workp. 15
  20. Acknowledgementsp. 15