https://lilianweng.github.io/posts/2024-02-05-human-data-quality/
The alignment stage of LLM post-training requires lots of human-generated data, and the quality is bounded on how good that data is. Lots of techniques here, with citations.
https://lilianweng.github.io/posts/2024-02-05-human-data-quality/
The alignment stage of LLM post-training requires lots of human-generated data, and the quality is bounded on how good that data is. Lots of techniques here, with citations.