基座模型安全产品 - AI创新业务
Core
Design and build safety capabilities for the Doubao base large language model, defining ideal model performance across normal, boundary, and high-risk scenarios to enhance safety and user experience.
Role type
Senior Product Manager, AI Safety & Content Governance
Builds
Content safety evaluation systems, automated risk detection tools, and safety-aligned training data assets for the Doubao model.
Domain
Generative AI / Large Language Models / Content Safety
Deliverable
production ML models
Required skills
Large model training methodology, safety strategy design, content governance, evaluation system design, data analysis, automated testing, cross-functional collaboration
Preferred skills
Risk classification, adversarial attack analysis, metric design, project management
Technologies
Large language models, automated evaluation tools, data labeling platforms
Responsibilities
Define model safety standards for various risk scenarios; build and maintain content safety evaluation frameworks; establish data standards for safety training and quality assessment; develop automated risk detection and evaluation tools; collaborate with engineering and data teams to close the loop on safety issues and validate improvements.
