Research Professional I, Computer Science (Temporary)

Implement the basic Monte Carlo Tree Search algorithm and evaluate it in publicly-available benchmark reinforcement learning environments such as OpenAI Gym. Implement basic imitation learning algorithms such as supervised finetuning (SFT) and evaluate them in benchmark reinforcement learning environments. Implement an reinforcement learning algorithm using the outcomes of tasks 1 and 2, and prepare evaluation reports for this algorithm in benchmark reinforcement learning environments.

Maintaining source code of the training pipelines, such as documentation of the source code, preparing usage manuals, and identification of library dependencies. Collaborate with internal departments to share information and improve data handling processes. Knowledge, Skills, and Abilities Skilled in maintaining code for libraries, APIs, deep learning, and reinforcement learning pipelines.

Ability to generate evaluation plots on the performance of algorithms, and visualize the reasoning traces of Monte-Carlo Tree Search agents Ability to translate technical concepts into clear documentation and collaborate across multidisciplinary teams.