2 months ago
Tokyo, JapanMid Level / Senior
Responsibilities
- Design and execute post-training strategies for language and multimodal models.
- Build and curate high-quality training data with a focus on Japanese-language capabilities.
- Develop evaluations to identify capability and reliability gaps.
- Conduct systematic error analysis to improve training methods and model behavior.
- Run controlled experiments and communicate clear recommendations.
- Develop reliable training and evaluation pipelines in collaboration with infrastructure teams.
- Contribute methods and findings to accelerate post-training work.
Requirements
- Hands-on experience with post-training modern language or multimodal models.
- Strong understanding of machine learning fundamentals and current post-training methods.
- Solid engineering skills and proficiency with the open-source ML ecosystem.
- Experience designing and running rigorous experiments, including systematic error analysis.
- Ability to turn research ideas into reliable implementations and measurable improvements.
- Proficiency in English for collaboration with global teams.
- Experience leveraging agents to enhance work.
Benefits
- High-impact model work improving foundation models in real products.
- Direct access to core teams for close collaboration.
- Ownership in a growing market with strong customer demand.
- Supportive culture prioritizing continuous learning and mentorship.
- Primarily remote work with flexible hours and unlimited paid time off.
- Competitive salary, equity in a unicorn-stage company, and standard benefits.
Tech Stack
Categories
Data Science
About Liquid AI
Liquid AI builds general-purpose AI systems that run efficiently from data center accelerators to on-device hardware, emphasizing low latency, memory efficiency, privacy, and reliability. The company partners with enterprises in consumer electronics, automotive, life sciences, and financial services to deploy and benchmark models for real-world workloads. Founded in 2023 out of MIT CSAIL and headquartered in Cambridge, Massachusetts, it is privately held.