[AI Research Div.] Foundation Model Evaluation Engineer - 독자 AI 파운데이션 모델 (2년 이상 / 인턴)
Core
Design and build evaluation benchmarks and pipelines to assess the capabilities of foundation models and frontier LLMs, ensuring data integrity and providing insights for model improvement.
Role type
Foundation Model Evaluation Engineer (Research)
Builds
Novel evaluation benchmarks, evaluation pipelines/platforms, and technical reports/papers.
Domain
AI/ML, NLP, Foundation Models
Deliverable
production ML models
Required skills
Deep learning research, AI model evaluation/analysis, benchmark design, data contamination analysis, experimental design, technical writing, rapid domain adaptation
Preferred skills
Top-tier conference publication experience, data collection/generation/curation for evaluation, benchmark/toolkit design, agentic flow design, large-scale evaluation infrastructure, extensive foundation model usage
Technologies
LLMs, Foundation Models, Evaluation Frameworks
Responsibilities
Define model capabilities and targets to guide development direction; design and build new benchmarks immune to data contamination; diagnose frontier models to support technical decisions; collaborate with data/model teams to identify weaknesses and design improvements; develop and manage large-scale evaluation pipelines; publish research at top-tier conferences (NeurIPS, ICLR, ICML, ACL, EMNLP).