Make LLMs Smarter, Safer, and More Useful | Research Engineer | Remote-First

• Reasoning optimization for coding, mathematics, logic, and conversation
• Hallucination reduction and model safety
• LLM evaluation metrics and benchmarks
• Synthetic-data generation and filtering pipelines
• Distributed training with PyTorch, DeepSpeed, or FSDP

• Compensation starts from 15M, even for junior-level researchers
• Remote-first, with full remote work available from anywhere in Japan
• Stock options
• Strong computing resources and long-term R&D funding
• Direct impact on the quality of a global consumer AI product

キャリア、ポジション等ご質問のある方はご連絡下さい。ぜひご連絡ください。