Agent Post-Training Context Research

Skills

About the Role

You will develop and test approaches for scaling compute spent on context in the frontier training stack. You will improve post-training systems, build evaluations and training environments, turn product signals into model improvements, and investigate model failures through experiments, analysis, and launch.

Requirements

• Machine learning
• Software engineering
• Systems
• Statistics
• LLMs
• RL
• RLHF/RLAIF
• Post-training
• Evaluations
• Graders
• Synthetic data
• Model training
• Coding agents
• Tool-using agents
• Production ML systems

Responsibilities

• Design and run experiments that improve scaling of compute on context
• Own improvements to the post-training stack, including RL, data pipelines, graders, reward signals, evals, diagnostics, and model-behavior analysis
• Build evaluations and environments that expose model failures
• Translate product signals into model improvements
• Work on training and alignment interventions
• Improve large-scale training and launch reliability, observability, reproducibility, cost, latency, and production readiness
• Debug model failures and turn them into hypotheses, experiments, and fixes

Benefits

• Medical, dental, and vision insurance with employer Health Savings Account contributions
• Pre-tax Health FSA, Dependent Care FSA, and commuter accounts
• 401(k) retirement plan with employer match
• Paid parental, medical, and caregiver leave
• Flexible PTO or up to 15 days of annual PTO, depending on exemption status
• 13+ paid holidays, coordinated office closures, and paid sick or safe time
• Mental health and wellness support
• Employer-paid basic life and disability coverage
• Annual learning and development stipend
• Daily office meals and eligible meal-delivery credits
• Relocation support for eligible employees
• Charitable donation matching and wellness stipends may be provided
• Equity
• Performance-related bonuses for eligible employees