UltraFeedback: Boosting Language Models with High-quality Feedback
OpenBMB 64K-prompt preference dataset where 17 LLMs respond and GPT-4 grades each response on instruction-following, truthfulness, honesty, and helpfulness.
- Publisher
- Tsinghua OpenBMB
- Year
- 2023
- Venue
- ICML
- Authors
- 12
- Hosting
- External sourcelicense unknown
Cite
Notes
Only stored in your browser.
Introduces 1 artifact - 1 tool
TL;DR
Semantic Scholar
This work validates the effectiveness of scaled AI feedback data in constructing strong open-source chat language models, serving as a solid foundation for future feedback learning research.
Artifacts
1Tools