SlopToPolish
收藏官方服务：
资源简介：
The edit model training dataset from [AI-Slop to AI-Polish? Aligning Language Models through Edit-Based Writing Rewards and Test-time Computation](https://arxiv.org/abs/2504.07532). Ideally, a model trained on this would be combined with the WQRM reward model from the same paper.
提供机构：
maas创建时间：
2026-03-25
