Shape the future of AI as a Principal Research Scientist at Autodesk, focusing on post-training alignment in a collaborative environment. Engage with innovative algorithms and real-world applications in architecture and engineering.
This senior role at Autodesk emphasizes deep expertise in reinforcement learning and post-training methods such as RLHF and PPO. You will be responsible for developing algorithms to enhance model alignment, reliability, and controllability while establishing evaluation systems. Collaborate closely with infrastructure teams and engage in rigorous model analysis to ensure high-quality outcomes.
Key Responsibilities:
• Develop algorithms for model reliability and alignment
• Design and run experiments for model behavior analysis
• Lead evaluation framework for reasoning and tool use
• Communicate technical risks and trade-offs to leadership
• Partner with teams to create scalable post-training workflows
Requirements:
• PhD or equival...