arXivDaily arXiv每日学术速递 周一至周五更新

大厂专区

Xiaohongshu(小红书)

2026-08-26 至 2026-08-26 共收录 2
2608.24168 2026-08-26 cs.CL cs.SD 新提交

FireRedAudio: A General-Purpose Audio Language Model with Decoupled Continuous Representations for Understanding and Generation

FireRedAudio:一种具有解耦连续表示的通用音频语言模型,用于理解与生成

Junjie Li, Xuelong Geng, Kun Xie, Feiyu Shen, Yichen Wu, Ziqi Dai, Yichen Han, Yan Jia, Kai Huang, Junjie Chen, Yixuan Li, Manzhen Wei, Fenglong Xie, Lei Xie, Xu Tang, Yao Hu

机构 * Xiaohongshu(小红书)

AI总结 FireRedAudio是首个公开的统一音频-语言模型,采用解耦连续表示,支持音频理解、多语种ASR、各类TTS及语音编辑,性能优于相关模型。

Comments 20 pages, 3 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2605.26958 2026-08-26 cs.CL cs.AI 版本更新

Tournament-GRPO: Group-Wise Tournament Rewards for Reinforcement Learning in Open-Ended Long-Form Generation

Tournament-GRPO:面向开放式长文本生成强化学习的群组锦标赛奖励

Zixuan Yang, Yiqun Chen, Wei Yang, Erhan Zhang, Zihan Shen, Xiaochi Wei, Yan Gao, Yi Wu, Yao Hu, Jiaxin Mao

机构 * Renmin University of China(中国人民大学) University of Southern California(南加州大学) Zhejiang University(浙江大学) Xiaohongshu Inc.(小红书公司)

AI总结 针对开放式长文本生成中缺乏可靠参考答案和自动评估指标的问题,提出Tournament-GRPO框架,通过同一查询生成结果间的多轮锦标赛比较将基于规则的LLM评判转化为相对奖励,在Deep Research Bench上取得4.52分提升。

Comments Accepted to EMNLP 2026 Main Conference

详情

展开后加载摘要…

URL PDF HTML 收藏