Applied Scientist, Wayve

• The ideal candidate has a deep understanding of reinforcement learning, machine learning, and behavioural modelling, combined with a drive to innovate in the autonomous driving space
• Proven expertise in reinforcement learning, including in areas like offline RL, reward modelling, RLHF, DPO, GPRO, as well as experience with machine learning
• Strong programming skills in Python and experience with machine learning libraries such as PyTorch
• Experience in working with simulation environments and real-world data for model validation and performance benchmarking
• Demonstrated ability to publish research and present findings to both technical and non-technical audiences at top tier conferences
• Excellent problem-solving skills and the ability to work independently as well as in a team environment
• Demonstrated ability to work collaboratively in a fast-paced, innovative, interdisciplinary team environment

Desirable

• We're looking for an experienced Applied Scientist with expertise in Reinforcement Learning and Reward Modelling to advance our training and evaluation frameworks contributing significantly to the creation of safe and reliable AI driving technology.
• In this role, you will be at the forefront of designing and optimizing reward and reinforcement learning models that are powerful and resource-efficient, tailored for the unique demands of embodied AI and autonomous systems. Your work will involve but not limited to:
• Design, develop, and refine reward models that align with safe and efficient driving objectives for autonomous vehicles
• Work closely with multidisciplinary teams to integrate reward models with real-world data and simulation frameworks
• Define a data strategy that includes efficient use of real and synthetic data, annotations, and active learning
• Design experiments to evaluate reward structures in diverse driving scenarios and identify areas for improvement
• Collaborate with world-class researchers and engineers to push the boundaries of AI, contributing significantly to the evolution of autonomous driving technology