腾讯地图-多模态VLM感知以及COT推理高级工程师
Tencent
腾讯地图-多模态VLM感知以及COT推理高级工程师
Job Description
腾讯地图-多模态VLM感知以及COT推理高级工程师
📋 Job Overview
The role focuses on leading the research and development of multimodal large language models (VLM) for digital maps at Tencent Maps, emphasizing breakthroughs in visual-language joint representation and cross-modal semantic alignment. It involves designing Chain-of-Thought (COT) inference frameworks to enhance decision-making in complex spatial scenarios and driving the engineering implementation of multimodal perception algorithms in advanced applications like autonomous driving lane-level navigation and SD maps. The position requires handling multimodal data processing, model training, optimization, and cross-team collaboration to align solutions with business needs.
📍 Location: Beijing, China
🏢 Business Unit: CSIG
📄 Full Description
2.设计并构建基于COT思维链（Chain-of-Thought）的推理框架，优化复杂空间场景下的决策逻辑生成能力；
3.推动多模态感知算法在自动驾驶车道级导航地图、标准地图（SD地图）等前沿场景中的工程化落地与性能优化；
4.负责多模态数据的处理、模型训练与调优，持续提升算法在复杂场景下的鲁棒性与泛化能力；
5.与跨团队协作，包括产品、工程和数据科学团队，确保算法方案与业务需求高效对齐并快速迭代。
🎯 Key Responsibilities
- Lead the algorithm R&D and innovation of multimodal large language models (VLM) in the digital map field, focusing on breakthroughs in visual-language joint representation, cross-modal semantic alignment, and other core technologies.
- Design and build a reasoning framework based on Chain-of-Thought (COT), optimizing decision logic generation capabilities in complex spatial scenarios.
- Promote the engineering implementation and performance optimization of multimodal perception algorithms in cutting-edge scenarios such as autonomous driving lane-level navigation maps and standard maps (SD maps).
- Handle multimodal data processing, model training, and tuning to continuously improve algorithm robustness and generalization in complex scenarios.
- Collaborate with cross-functional teams, including product, engineering, and data science teams, to ensure efficient alignment of algorithm solutions with business needs and rapid iteration.
Locations
- Beijing, China
Salary
400,000 - 800,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Lead the algorithm R&D and innovation of multimodal large language models (VLM) in the digital map field, focusing on breakthroughs in visual-language joint representation, cross-modal semantic alignment, and other core technologies.
- Design and build a reasoning framework based on Chain-of-Thought (COT), optimizing decision logic generation capabilities in complex spatial scenarios.
- Promote the engineering implementation and performance optimization of multimodal perception algorithms in cutting-edge scenarios such as autonomous driving lane-level navigation maps and standard maps (SD maps).
- Handle multimodal data processing, model training, and tuning to continuously improve algorithm robustness and generalization in complex scenarios.
- Collaborate with cross-functional teams, including product, engineering, and data science teams, to ensure efficient alignment of algorithm solutions with business needs and rapid iteration.
Tags & Categories
腾讯地图-多模态VLM感知以及COT推理高级工程师
Tencent
腾讯地图-多模态VLM感知以及COT推理高级工程师
Job Description
腾讯地图-多模态VLM感知以及COT推理高级工程师
📋 Job Overview
The role focuses on leading the research and development of multimodal large language models (VLM) for digital maps at Tencent Maps, emphasizing breakthroughs in visual-language joint representation and cross-modal semantic alignment. It involves designing Chain-of-Thought (COT) inference frameworks to enhance decision-making in complex spatial scenarios and driving the engineering implementation of multimodal perception algorithms in advanced applications like autonomous driving lane-level navigation and SD maps. The position requires handling multimodal data processing, model training, optimization, and cross-team collaboration to align solutions with business needs.
📍 Location: Beijing, China
🏢 Business Unit: CSIG
📄 Full Description
2.设计并构建基于COT思维链（Chain-of-Thought）的推理框架，优化复杂空间场景下的决策逻辑生成能力；
3.推动多模态感知算法在自动驾驶车道级导航地图、标准地图（SD地图）等前沿场景中的工程化落地与性能优化；
4.负责多模态数据的处理、模型训练与调优，持续提升算法在复杂场景下的鲁棒性与泛化能力；
5.与跨团队协作，包括产品、工程和数据科学团队，确保算法方案与业务需求高效对齐并快速迭代。
🎯 Key Responsibilities
- Lead the algorithm R&D and innovation of multimodal large language models (VLM) in the digital map field, focusing on breakthroughs in visual-language joint representation, cross-modal semantic alignment, and other core technologies.
- Design and build a reasoning framework based on Chain-of-Thought (COT), optimizing decision logic generation capabilities in complex spatial scenarios.
- Promote the engineering implementation and performance optimization of multimodal perception algorithms in cutting-edge scenarios such as autonomous driving lane-level navigation maps and standard maps (SD maps).
- Handle multimodal data processing, model training, and tuning to continuously improve algorithm robustness and generalization in complex scenarios.
- Collaborate with cross-functional teams, including product, engineering, and data science teams, to ensure efficient alignment of algorithm solutions with business needs and rapid iteration.
Locations
- Beijing, China
Salary
400,000 - 800,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Lead the algorithm R&D and innovation of multimodal large language models (VLM) in the digital map field, focusing on breakthroughs in visual-language joint representation, cross-modal semantic alignment, and other core technologies.
- Design and build a reasoning framework based on Chain-of-Thought (COT), optimizing decision logic generation capabilities in complex spatial scenarios.
- Promote the engineering implementation and performance optimization of multimodal perception algorithms in cutting-edge scenarios such as autonomous driving lane-level navigation maps and standard maps (SD maps).
- Handle multimodal data processing, model training, and tuning to continuously improve algorithm robustness and generalization in complex scenarios.
- Collaborate with cross-functional teams, including product, engineering, and data science teams, to ensure efficient alignment of algorithm solutions with business needs and rapid iteration.
Tags & Categories