大模型策略产品经理-安全方向（J92070）
Core
Design and execute comprehensive evaluation frameworks (manual and automated) for large language model (LLM) safety, collaborating with algorithm teams to optimize RL training strategies and enhance model performance in weak areas.
Role type
Senior Product Manager (LLM Safety & Evaluation)
Builds
Automated LLM evaluation benchmarks and RL training environments
Domain
Artificial Intelligence / Large Language Models / Cybersecurity
Deliverable
production ML models
Required skills
LLM safety evaluation design, Reinforcement Learning strategy, Python data analysis, System architecture design, AI tool orchestration
Preferred skills
Deep understanding of LLM architecture and training pipelines, Experience with complex scenario benchmarking
Technologies
Python, RL training environments, AI tools
Responsibilities
Build automated LLM safety evaluation benchmarks covering complex scenarios; Collaborate with algorithm teams to design RL training environments and optimize training strategies; Collect and analyze data to support benchmark optimization and model tuning; Track industry trends and competitor analysis to drive product strategy innovation.
Seniority
Senior, hands-on IC