Paper Archive

输入关键词搜索 102 篇论文

tracked author

Bo Zheng

Recurring Qwen author across Qwen2.5, Qwen3 and Bebop; RollArt adds Alibaba ROLL / agentic RL infrastructure linkage. Scholar/Hugging Face and Qwen technical reports provide stable Qwen Team linkage, but public personal profile evidence is sparse. No high-confidence personal X account was found.

3 archived notes X: not-found Hugging Face

Related Notes

按论文归档时间排序，展示该作者在本站已经出现的材料。

Jun 30, 2026

RollArt: Disaggregated Multi Task Agentic RL Training at Scale

RollArt 用静态 task domain affinity 声明把训练、prefill heavy rollout、decode heavy rollout、环境与 reward 映射到 H800、H20、CPU/Kubernetes 和 serverless 资源池，再用轨迹级状态机、起始版本年龄上限 $\alpha$ 与 Mooncake weight movement 协调跨池异步；OSDI camera ready 支持...

2512.22560-rollart-disaggregated-agentic-rl-training RL InfrastructureRollout OptimizationDistributed Training
Jun 23, 2026

Qwen2.5 Technical Report

Qwen2.5 展示了一条面向通用、代码、数学、结构化数据和长上下文场景的 LLM (Large Language Model，大语言模型) 工程路线：用 18T token 预训练、统一 tokenizer、百万级 SFT (Supervised Fine Tuning，监督微调)、DPO (Direct Preference Optimization，直接偏好优化)、GRPO (Group Relative Policy Opti...

2412.15115-qwen2-5-technical-report Reasoning RLLong ContextCoding Agent
Jun 10, 2026

Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling

Bebop 的核心判断是：MTP 在 RL rollout 中失速的主因主要来自 policy entropy fluctuation，frozen MTP head 与更新后 policy 的权重漂移影响较小；因此有效方案是把 acceptance method 换成 probabilistic rejection sampling，并用 end to end TV loss 在 RL 前训练 MTP heads，使 draft t...

2606.12370-bebop-mtp-rejection-sampling-rl-training Rollout OptimizationSpeculative DecodingMulti-Token Prediction