ByteDance
Posted 4w ago

大模型算法研究员 - 小荷健康

ByteDance
Beijing, Beijing, China
OnsiteFull Time
Responsibilities
  • training models
  • designing experiments
  • collaborating cross-team
Requirements
  • Master's degree (2027) in AI/computer science/math/bioinformatics/statistics preferred
  • Strong coding
  • LLM training and alignment knowledge (pretraining/SFT/RL/RLHF/DPO/PPO)
  • Experiment design and research skills
  • Interest in medical AI
Technical tools mentioned
MidTrainSFTRLRLHFDPOPPOReward ModelingAI InfraHarnessLoopGraph Engineering

Job description

团队介绍:小荷健康是字节跳动旗下的医疗健康品牌,我们致力于利用大模型技术,探索医疗健康服务的新范式,孵化了“小荷AI医生”app,为用户提供专业的健康咨询服务。小荷健康拥有专业的医学和互联网产研团队,认真探索AI医疗。

1、参与医疗基座模型训练与优化,涵盖数据构建、MidTrain、SFT、RL、偏好优化等关键环节;结合AI Infra、Harness/Loop/Graph Engineering,完善训练、推理、评测与自动化实验体系,提升研发效率与训练效果;
2、围绕医疗复杂任务开展算法研究,重点突破Long-Horizon场景下的规划、推理、执行、恢复与自我修正能力;
3、参与医疗Agent任务模式建设,优化任务编排、工具调用、状态跟踪、记忆管理、反思纠错等核心机制;
4、面向生物信息、EHR仿真、临床对话、医疗问答等任务设计实验、构建评测体系、分析误差并迭代模型能力;
5、跨团队协作推动模型能力在真实业务场景落地,对最终效果负责。

1、2027届获得硕士及以上学位,人工智能、计算机、数学、生物信息学、统计等相关专业优先;
2、具备良好的AI-Native/Vibe coding能力,能够高效借助AI工具完成实验实现、问题调试和迭代验证;有ICML、NeurIPS、ICLR、ACL、EMNLP、CVPR、MICCAI等顶会相关论文者优先;
3、对大模型训练、对齐与推理有系统理解,了解预训练、SFT、RL、偏好优化中的一种或多种;有大模型预训练、SFT、RLHF/DPO/PPO、Reward Modeling经验者优先;
4、对医疗AI、长程任务、Agent、模型训练与优化有强烈兴趣,愿意深入难题并持续做高质量研究;有医疗大模型、临床NLP、生物信息建模、EHR任务、医疗知识建模经验者优先;有Agent训练经验者优先,尤其是在Long-Horizon Task、Workflow Graph、Memory/Retrieval、自反思、自进化等方向有实践经验;
5、具备较强的实验设计、问题分析、论文阅读和结果归因能力,能够从数据和现象中定位本质问题;有业界通用基模后训练经验者优先,尤其是在Reasoning、Tool-Use、Deep Research、Agent能力提升等方向有训练或优化经验,或具备多模态感知、推理、偏好优化等相关实践;
6、具备良好的协作意识和自驱力,能够独立推进课题并持续产出;有高质量开源项目、工业界实习或医疗AI项目落地经验者优先。

About ByteDance

Developing AI-driven content platforms and mobile applications.

Similar jobs

Algorithm Researcher roles near Beijing, Beijing
3d
Save
Mark Applied
Hide
算法研究员(蛋白质结构生成方向)
Beijing, Beijing, China
OnsiteFull Time
北京科学智能研究院
北京科学智能研究院: A research institute focused on AI for Science innovation.
PhD or higher in progress preferred in computer science, software engineering, artificial intelligence, bioinformatics, or related fields; Python and deep learning framework experience required.
Python, PyTorch, JAX, AlphaFold3
6d
Save
Mark Applied
Hide
ATH MaaS大模型&智能体前沿算法研究-阿里星/A Star
Beijing or Hangzhou
RemoteFull Time
Alibaba Group
Alibaba GroupNYSE: BABA: Global technology specializing in e-commerce and cloud computing.
Excellent master's or doctoral background in computer science, AI, software engineering, mathematics, or related fields; strong deep learning, distributed training, model optimization, research, and coding experience.
Alibaba Cloud, Agentic RL, Agentic Framework, AI搜索, AI客服, AI视频生成, NeurIPS, ICML, ICLR, ACL, EMNLP, ICASSP, AutoResearch
6d
Save
Mark Applied
Hide
ATH MaaS大模型&智能体前沿算法研究-阿里星/A Star
Beijing or Hangzhou
OnsiteFull Time
Alibaba
AlibabaNYSE: BABA: Provides online marketplaces, cloud computing, and digital payment services.
Excellent master's or doctoral background in computer science, artificial intelligence, software engineering, mathematics, or related fields; strong deep learning, distributed training, inference, coding, and large-model research experience.
Alibaba Cloud, Agentic RL, Agentic Framework, AI Search, AI Customer Service, AI Video Generation, NeurIPS, ICML, ICLR, ACL, EMNLP, ICASSP, AutoResearch
1w
Save
Mark Applied
Hide
算法-其他方向
Shenzhen or Beijing or Shanghai
OnsiteFull Time
Tencent
TencentHong Kong Stock Exchange: 0700: Multinational technology conglomerate providing internet and entertainment services.
PhD or excellent master's degree in computing, engineering, AI, mathematics, physics, or related fields; proficiency in at least one programming language; research or project experience preferred.
Java, C, C++, C#, Python
1w
Save
Mark Applied
Hide
算法-其他方向
Shenzhen or Beijing or Shanghai
OnsiteFull Time
Tencent
TencentHKEX: 0700: Provides integrated internet services, digital entertainment, and cloud technology.
PhD or outstanding master's degree in a relevant technical field, proficiency in at least one programming language, and research or project experience in a relevant direction.
Java, C, C++, C#, Python
2w
Save
Mark Applied
Hide
视频生成基座模型算法专家
Shanghai or Beijing
OnsiteFull Time
Resonate
Resonate: AI-powered consumer intelligence and predictive data analytics platform.
Master's degree or higher in CS/AI, deep ML/DL theory, first-principles understanding of diffusion or autoregressive models, experience leading large-scale generative model training and alignment (SFT/RLHF).
Diffusion, Transformer, SFT, RLHF
3w
Save
Mark Applied
Hide
Super Sparks-校招-座舱具身智能体算法研究员
Beijing or Shanghai
OnsiteFull Time
Nio
NioNYSE: NIO: Designs and manufactures smart premium electric vehicles.
PhD required. Expert in statistical learning and deep learning (Transformer, Diffusion, VAE, Flow Matching). Familiar with reinforcement/behavioral policy learning (RLHF, PPO, DPO, GRPO). Experience training large models and multimodal/embodied interaction research.
Transformer, Diffusion, VAE, Flow Matching, RLHF, PPO, DPO, GRPO, TTS
3w
Save
Mark Applied
Hide
顶尖实习-具身大模型算法研究员-机器人事业部-实习1
Beijing, Beijing, China
OnsiteInternship
Xiaomi
XiaomiHKEX: 1810: Design and manufacturing of consumer electronics and smart devices.
PhD in CV/ML/Robotics/AI or research-oriented Master's; experience in VLM/multimodal/3D/Embodied Agent/RL; strong publication or impactful OSS background preferred; proficient in PyTorch and familiar with Transformer/Diffusion/CLIP/SigLIP/Qwen-VL.
PyTorch, Transformer, Diffusion, CLIP, SigLIP, Qwen-VL