混元大模型评测算法研究员（北京）
Tencent
混元大模型评测算法研究员（北京）
Job Description
混元大模型评测算法研究员（北京）
📋 Job Overview
The role of Algorithm Researcher for Hunyuan Large Model Evaluation at Tencent in Beijing focuses on planning, implementing, and building platform capabilities for evaluating and applying general AI large models. Key responsibilities include constructing evaluation benchmarks for multimodal capabilities, developing automated data production pipelines, and implementing automated evaluation and attribution analysis mechanisms. This position drives the advancement of comprehensive, accurate, and evolving evaluation frameworks for large models across text, vision, audio, video, and other modalities.
📍 Location: Shenzhen, China
🏢 Business Unit: TEG
📄 Full Description
1.通用AI大模型评测基准的构建：建立覆盖文生文、多模态理解、多模态（音视频/3D/图/视频生成）生成等多模态的评测基准，设计全面、准确的多维度指标，构建自动化评测工具链，并随着模型能力的演进持续探索全模态的评测基准；
2.评测数据的自动化生产能力构建：基于数据泛化等能力，构建领域增强型评测数据集生产链路，支持多模态场景的自动化数据扩增与效果验证；
3.自动化评测与归因分析：探索并实现各个模态大模型的自动化评测与模型缺陷归因机制。
🎯 Key Responsibilities
- Build evaluation benchmarks for general AI large models, covering text generation, multimodal understanding, and multimodal generation (audio/video/3D/image/video), including designing multi-dimensional indicators, creating automated evaluation toolchains, and continuously exploring full-modal benchmarks as model capabilities evolve.
- Construct automated production capabilities for evaluation data, leveraging data generalization to build domain-enhanced evaluation dataset production pipelines that support automated data augmentation and effect validation in multimodal scenarios.
- Explore and implement automated evaluation and attribution analysis for large models across various modalities, including mechanisms for model defect attribution.
Locations
- Shenzhen, China
Salary
300,000 - 600,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Build evaluation benchmarks for general AI large models, covering text generation, multimodal understanding, and multimodal generation (audio/video/3D/image/video), including designing multi-dimensional indicators, creating automated evaluation toolchains, and continuously exploring full-modal benchmarks as model capabilities evolve.
- Construct automated production capabilities for evaluation data, leveraging data generalization to build domain-enhanced evaluation dataset production pipelines that support automated data augmentation and effect validation in multimodal scenarios.
- Explore and implement automated evaluation and attribution analysis for large models across various modalities, including mechanisms for model defect attribution.
Tags & Categories
Answer 10 quick questions to check your fit for 混元大模型评测算法研究员（北京） @ Tencent.
Related Books and Jobs
混元大模型评测算法研究员（北京）
Tencent
混元大模型评测算法研究员（北京）
Job Description
混元大模型评测算法研究员（北京）
📋 Job Overview
The role of Algorithm Researcher for Hunyuan Large Model Evaluation at Tencent in Beijing focuses on planning, implementing, and building platform capabilities for evaluating and applying general AI large models. Key responsibilities include constructing evaluation benchmarks for multimodal capabilities, developing automated data production pipelines, and implementing automated evaluation and attribution analysis mechanisms. This position drives the advancement of comprehensive, accurate, and evolving evaluation frameworks for large models across text, vision, audio, video, and other modalities.
📍 Location: Shenzhen, China
🏢 Business Unit: TEG
📄 Full Description
1.通用AI大模型评测基准的构建：建立覆盖文生文、多模态理解、多模态（音视频/3D/图/视频生成）生成等多模态的评测基准，设计全面、准确的多维度指标，构建自动化评测工具链，并随着模型能力的演进持续探索全模态的评测基准；
2.评测数据的自动化生产能力构建：基于数据泛化等能力，构建领域增强型评测数据集生产链路，支持多模态场景的自动化数据扩增与效果验证；
3.自动化评测与归因分析：探索并实现各个模态大模型的自动化评测与模型缺陷归因机制。
🎯 Key Responsibilities
- Build evaluation benchmarks for general AI large models, covering text generation, multimodal understanding, and multimodal generation (audio/video/3D/image/video), including designing multi-dimensional indicators, creating automated evaluation toolchains, and continuously exploring full-modal benchmarks as model capabilities evolve.
- Construct automated production capabilities for evaluation data, leveraging data generalization to build domain-enhanced evaluation dataset production pipelines that support automated data augmentation and effect validation in multimodal scenarios.
- Explore and implement automated evaluation and attribution analysis for large models across various modalities, including mechanisms for model defect attribution.
Locations
- Shenzhen, China
Salary
300,000 - 600,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Build evaluation benchmarks for general AI large models, covering text generation, multimodal understanding, and multimodal generation (audio/video/3D/image/video), including designing multi-dimensional indicators, creating automated evaluation toolchains, and continuously exploring full-modal benchmarks as model capabilities evolve.
- Construct automated production capabilities for evaluation data, leveraging data generalization to build domain-enhanced evaluation dataset production pipelines that support automated data augmentation and effect validation in multimodal scenarios.
- Explore and implement automated evaluation and attribution analysis for large models across various modalities, including mechanisms for model defect attribution.
Tags & Categories
Answer 10 quick questions to check your fit for 混元大模型评测算法研究员（北京） @ Tencent.
Related Books and Jobs