[Skip to main content](#main-content)

[OpenNash Jobs](https://jobs.opennash.com/)
[Hiring report](/reports/hiring/) · [All Tencent Holdings jobs](https://jobs.opennash.com/companies/tencent-holdings/) · [Search jobs](/search/)

Tencent Holdings · Beijing

# 混元vlm算法研究员-后训练/基座能力提升方向（深圳/北京/上海）

Beijing 

Listed on Tencent Holdings’s official careers site. Listed as of October 19, 2026. Check that it is still open and apply there.
[Apply on Tencent Holdings’s careers site →](https://careers.tencent.com/en-us/jobdesc.html?postId=2074683054592536576)

[OpenNash Jobs](https://jobs.opennash.com/) / [Tencent Holdings](https://jobs.opennash.com/companies/tencent-holdings/) / 混元vlm算法研究员-后训练/基座能力提升方向（深圳/北京/上海）

## Job description

### 主导后训练研发体系
- 负责多模态大语言模型（VLM）的后训练（Post-training）全流程算法研发，深入探索 SFT 及强化学习对齐算法（如 RLHF、PPO、GRPO 等），持续逼近并提升视觉基座模型的能力上限；

### 攻坚视觉基础感知
- 致力于优化模型底层的视觉感知、细粒度理解与空间感知能力。前瞻性地探索包括“带图思考”在内的 Visual Agent 解决方案，赋予模型更深层次的物理世界理解与执行能力；

### 提高大模型可靠性
- 致力于大幅度降低多模态大模型幻觉。深入探究幻觉产生的本质，建设完善的幻觉检测链路，实施精细化的数据清洗策略，并主导研发基于强化学习的幻觉消除（Hallucination Removal）方案；

### 赋能用户交互场景
- 面向真实用户场景进行模型对齐与调优，强化模型在多轮图文交互中的逻辑推理、指令遵循（Instruction Following）以及细节感知能力。搭建线上 Badcase 监控与数据飞轮反馈闭环，持续打磨并打造极致的交互体验。 Requirements
- 计算机科学/软件工程/人工智能等相关专业硕士及以上学历；
- 深入理解 Transformer 架构及大语言模型训练原理，在 LLM Alignment、RLHF、Reward Modeling、个性化大模型（Personalized LLM）、记忆机制（Memory/RAG）等后训练领域之一有深入的研究和实践经验；
- 在用户画像建模、推荐系统与大模型结合、或超长上下文/长期记忆模型优化方面有丰富实战经验者优先；
- 具备扎实的算法基础和工程实现能力，熟练掌握 Python，熟悉 PyTorch/TensorFlow 等深度学习框架；
- 拥有分布式训练实战经验，熟悉 Megatron-LM, DeepSpeed, vLLM 等大规模训练及推理框架，有百亿/千亿参数模型训练或调优经验者优先；
- 具备优秀的科研素养，有高质量论文发表（NeurIPS, ICLR, ICML, ACL, EMNLP 等）或在开源社区（如 HuggingFace）有高影响力项目贡献者优先；
- 具备强烈的技术热情和自驱力，善于分析和解决复杂问题，拥有良好的团队协作和沟通能力。

## At a glance
- Posted by the employer: Not stated
- First seen by OpenNash: Jul 19, 2026
- Last checked: Oct 19, 2026
- Listed for: 92 days
- Location: Beijing[Apply on Tencent Holdings’s careers site →](https://careers.tencent.com/en-us/jobdesc.html?postId=2074683054592536576)

## More jobs at Tencent Holdings

### [3D动作游戏《狩》-后台开发-UE5 DS后台开发](https://jobs.opennash.com/companies/tencent-holdings/jobs/6d535d770b7d94786b5e/)

Beijing

### [AI infra研发工程师(北京/深圳)](https://jobs.opennash.com/companies/tencent-holdings/jobs/eda44bea14836b2b9952/)

Beijing

### [AI ⽹络集合通信库运营⼯程师](https://jobs.opennash.com/companies/tencent-holdings/jobs/59f91606e7bdc470bf0b/)

Beijing

### [Aigc内容检测算法工程师（图像与视频方向）](https://jobs.opennash.com/companies/tencent-holdings/jobs/ca1cf721bb29a8df9bdf/)

Beijing

### [AI智能体评测高级工程师](https://jobs.opennash.com/companies/tencent-holdings/jobs/9312d7b2c30640f3f245/)

Beijing

## Similar jobs at other companies

### [【北斗】大模型训推引擎工程师（后训练/推理方向）](https://jobs.opennash.com/companies/meituan/jobs/1077b1379c49f4c43413/)

Meituan · 北京市 +1 more

### [MiMo算法研究员-后训练](https://jobs.opennash.com/companies/xiaomi/jobs/3908e9a1977af58b02f7/)

Xiaomi · 北京

OpenNash lists jobs from employers’ own careers sites and does not handle applications.
Dates come from the employer where stated; “first seen” and “last checked” are our own observations.

OpenNash Jobs · Official listings with dates and links to the employer. [Weekly hiring report](/reports/hiring/) · [Methodology](/methodology/)