腾讯混元大语言模型后训练算法工程师(深圳/北京/上海)
社招全职3年以上混元-模型算法技术地点:深圳状态:招聘
任职要求
1.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; 2.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; 3.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; 4.具备扎实的算法基础和工程实现能力,熟练掌握 Pytho…
登录查看完整任职要求
微信扫码,1秒登录
工作职责
1.负责大语言模型后训练(Post-Training)阶段的核心技术研发,构建和优化高质量的奖励系统(Reward System),通过Reward Modeling (RM) 和强化学习(RL)算法持续提升模型在复杂指令遵循、逻辑推理及价值观对齐方面的能力; 2.深入研究和优化 RLHF 等后训练算法,提升模型训练的稳定性和最终效果; 3.负责大模型个性化(Personalization)与长期记忆(Memory)机制的算法研发,构建精准的“千人千面”用户建模体系,探索模型如何理解、提取、记忆并动态适应不同用户的长期偏好,持续提升个性化交互体验; 4.负责后训练阶段的数据合成与管理,设计高效的数据飞轮机制,利用SFT、Self-Instruct等技术合成高质量训练数据,并负责建立从用户多维反馈(User Feedback)到模型迭代的闭环信号建模体系; 5.负责后训练模型的全维度评测与分析,制定科学的评价指标,跟进前沿技术动态,将最新研究成果快速转化为业务价值。
包括英文材料
学历+
Transformer+
https://huggingface.co/learn/llm-course/en/chapter1/4
Breaking down how Large Language Models work, visualizing how data flows through.
https://poloclub.github.io/transformer-explainer/
An interactive visualization tool showing you how transformer models work in large language models (LLM) like GPT.
https://www.youtube.com/watch?v=wjZofJX0v4M
Breaking down how Large Language Models work, visualizing how data flows through.
大模型+
https://www.youtube.com/watch?v=xZDB1naRUlk
You will build projects with LLMs that will enable you to create dynamic interfaces, interact with vast amounts of text data, and even empower LLMs with the capability to browse the internet for research papers.
https://www.youtube.com/watch?v=zjkBMFhNj_g
RLHF+
[英文] What is RLHF?
https://aws.amazon.com/what-is/reinforcement-learning-from-human-feedback/
Reinforcement learning from human feedback (RLHF) is a machine learning (ML) technique that uses human feedback to optimize ML models to self-learn more efficiently.
https://www.ibm.com/think/topics/rlhf
Reinforcement learning from human feedback (RLHF) is a machine learning technique in which a “reward model” is trained with direct human feedback, then used to optimize the performance of an artificial intelligence agent through reinforcement learning.
RAG+
https://www.youtube.com/watch?v=sVcwVQRHIc8
Learn how to implement RAG (Retrieval Augmented Generation) from scratch, straight from a LangChain software engineer.
推荐系统+
[英文] Recommender Systems
https://www.d2l.ai/chapter_recommender-systems/index.html
Recommender systems are widely employed in industry and are ubiquitous in our daily lives.
算法+
https://roadmap.sh/datastructures-and-algorithms
Step by step guide to learn Data Structures and Algorithms in 2025
https://www.hellointerview.com/learn/code
A visual guide to the most important patterns and approaches for the coding interview.
https://www.w3schools.com/dsa/
Python+
https://liaoxuefeng.com/books/python/introduction/index.html
中文,免费,零起点,完整示例,基于最新的Python 3版本。
https://www.learnpython.org/
a free interactive Python tutorial for people who want to learn Python, fast.
https://www.youtube.com/watch?v=K5KVEU3aaeQ
Master Python from scratch 🚀 No fluff—just clear, practical coding skills to kickstart your journey!
https://www.youtube.com/watch?v=rfscVS0vtbw
This course will give you a full introduction into all of the core concepts in python.
PyTorch+
https://datawhalechina.github.io/thorough-pytorch/
PyTorch是利用深度学习进行数据科学研究的重要工具,在灵活性、可读性和性能上都具备相当的优势,近年来已成为学术界实现深度学习算法最常用的框架。
https://www.youtube.com/watch?v=V_xro1bcAuA
Learn PyTorch for deep learning in this comprehensive course for beginners. PyTorch is a machine learning framework written in Python.
还有更多 •••
相关职位
社招1年以上混元-模型算法技
1.专业领域智能体架构升级:负责面向金融、法务、医疗等专业领域的大模型智能体体系建设,利用 Agentic RAG 架构实现从基础检索问答到深度分析、专业研判与复杂任务执行的演进; 2.深度推理与规划:针对复杂专业任务指令,设计并实现基于推理类模型的任务拆解、多步规划(Planning)与流程编排策略,提升系统处理专业分析、跨文档归纳、决策支持等复杂问题的能力; 3.事实核查与结果可靠性:建立可靠的事实核查(Fact-checking)与结果校验机制,通过证据溯源、多源信息比对、结构化校验等手段,解决大模型在专业场景中的幻觉与失真问题,确保生成结果的真实性、严谨性与可追溯性; 4.前沿技术转化:探索推理类模型在专业领域 Agent 的落地,包括基于过程监督的思维链(CoT)优化、面向专业任务反馈的强化学习(RL)策略,以及搜索、数据分析、文件处理等多工具协同能力的持续演进。
更新于 2026-07-21北京
社招1年以上混元-模型算法技
1.深度参与混元大模型后训练(包括 SFT、RL )及模型合版的研发和相关算法策略研究,提升模型泛化性、能力边界和上限; 2.后训练策略和生产范式的研究,探索更加敏捷、普适和可扩展的训练范式,提升训练和生产效率; 3.相关前沿后训练技术方向探索,包括但不限于 Reasoning/Agentic 能力增强和 Scaling, RL/OPD 等收敛性和可扩展性研究; 4.将研究成果转化为实际的模型能力或生产效率,并撰写高水平技术报告或论文,提升技术影响力。
更新于 2026-04-16北京