Researcher, Post Training

📋 Description

• Own research initiatives to improve multimodal model alignment
• Develop post-training methods and evaluation frameworks
• Partner with research, product, and platform teams to define best practices
• Implement, debug, and scale experimental ML systems
• Translate research into production-ready systems for improved alignment

🎯 Requirements

• Deep knowledge of preference optimization and RLHF methods
• Experience designing evaluations/metrics for generative or multimodal models
• Strong engineering and debugging skills for complex ML systems
• Ability to trace and diagnose model performance across training/evaluation

🎁 Benefits

• In-office policy: offices in SF, London, Bangalore
• Visa sponsorship support and case-by-case basis
• Health Insurance: medical, dental, vision for you and family
• Parental Leave: 9 weeks paternity; 12 weeks maternity
• 401(k) plan
• Commuter Allowance