happynew111/haotian_data-GPS-AR-Lopti-master
Do Not Let Low-Probability Tokens Over-Dominate in RL for LLMs Authors: Zhihe Yang$^{1}$, Xufang Luo$^{2*}$, Zilong Wang$^2$, Dongqi Han$^2$, Zhiyuan He$^2$, Dongsheng Li$^2$, Yunjian Xu$^{1*}$, ($^*$ for corresponding authors) The Chinese University of Hong Kong, Hong Kong SAR, China Microsoft Research Asia, Shanghai, China Introduction In this study, we identify a critical yet underexplored issue in RL training: low-probability tokens… See the full description on the dataset page: https://huggingface.co/datasets/happynew111/haotian_data-GPS-AR-Lopti-master.
| Author / organization | happynew111 |
|---|---|
| Downloads | 236 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 10 October 2025 |
| Last updated at source | 10 October 2025 |
Popularity, last 3 days
236 now (-8 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 16 Oct 2026 Description: Do Not Let Low-Probability Tokens Over-Dominate in RL for LLMs Authors: Zhihe Yang$^{1}$, Xufang Luo$^{2*}$, Zilong Wang$^2$, Dongqi Han$^2$, Zhiyuan He$^2$, Dongsheng Li$^2$, Yunjian Xu$^{1*}$, ($^*$ for corresponding authors) The Chinese → Do Not Let Low-Probability Tokens Over-Dominate in RL for LLMs Authors: Zhihe Yang$^{1}$, Xufang Luo$^{2*}$, Zilong Wang$^2$, Dongqi Han$^2$, Zhiyuan He$^2$, Dongsheng Li$^2$, Yunjian Xu$^{1*}$, ($^*$ for corresponding authors) The Chinese
Changes detected by Silytics between refreshes of the source.
More from happynew111
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.