大模型安全算法研究员（J106761）
Core
Design and build large model training schemes for vulnerability mining and exploitation in the cyber direction, including corpus construction and security verifier design.
Role type
Senior IC large model security algorithm researcher
Builds
Cyber security evaluation benchmarks and training datasets for vulnerability mining/exploitation
Domain
AI for Security / Cybersecurity
Deliverable
production ML models
Required skills
Large model training, vulnerability mining, security analysis, corpus construction, security verifier design, data distillation, adversarial attack/defense
Preferred skills
Publications in top AI conferences/journals or security Big 4
Technologies
Large language models, security frameworks
Responsibilities
Construct model network security capability evaluation systems and benchmark test sets; Design and build large model training schemes for vulnerability mining and exploitation; Follow up on frontier data distillation and security attack/defense technologies
Seniority
Senior, hands-on IC
