Tencent
Posted 2w ago

微信搜索-Agent算法专家

Tencent
Beijing, China
OnsiteFull Time
Responsibilities
  • optimizing agents
  • researching methods
  • exploring solutions
Requirements
  • Strong RL foundation with PPO/GRPO/DPO experience
  • Practical RLHF experience on large models
  • Familiarity with chain-of-thought and tool learning
  • Distributed training experience, and publications in top ML conferences preferred
Technical tools mentioned
Mid-TrainSFTGRMPRMRLVRPPOGRPODPORLHF

Job description

Responsibilities

1.参与微信搜索Agent能力优化,包括Search Agent(DeepSearch/DeepResearch)和真实世界复杂任务上的Agentic能力;

2.跟进前沿技术:Mid-Train、SFT、GRM、PRM、RLVR、Agentic RL、Agent自进化、Context管理/Memory等;

3.探索稳定高效的Agentic RL方案,探索下一代大模型结合Agent结合搜索的技术和产品范式。

Requirements

1.算法功底:精通PPO、GRPO、DPO等强化学习算法,有大模型RLHF实战经验;

2.专业深度:深刻理解CoT、自反思及工具学习,熟悉分布式训练框架;

3.学术背景:在NeurIPS、ICLR、ICML等顶会以一作发表过RL或Agent相关论文优先。

About Tencent

Provides integrated internet services, digital entertainment, and cloud technology.

Year founded
1998
Employees
112771
Organization type
Public
Headquarters
CN

Similar jobs

Agent Algorithm Specialist roles in Beijing
7mo
Save
Mark Applied
Hide
Agent算法专家-Commercial AI
Beijing, Beijing, China
OnsiteFull Time
ByteDance
ByteDance: Developing AI-driven content platforms and mobile applications.
Bachelor's degree in AI/computer science/automation/mathematics, strong ML/CV/NLP/RL and agent research background, excellent coding in C/C++ or Python, publication or large-project experience preferred.
C/C++, Python