R&D Programmer Assistant - Learning Bots That Adapt To New Tasks - LA Forge

Job Description

R&D Programmer Assistant - Learning bots that adapt to new tasks – La Forge (6 months internship)

Ubisoft Annecy

Founded in 1996, Ubisoft Annecy has developed many sagas, especially on Splinter Cell, Assassin's Creed®, while focusing on multiplayer online and social aspects, and also on concrete game mandates such as Tom Clancy's: The Division 2® and Ghost Recon Wildlands. We are now proud to have developed our own IP, Steep and more recently Riders Republic!
On top of our AAA production, we are working on innovation with a technological expertise hub around the Snowdrop engine pipeline, and Technology Group in collaboration with Ubisoft Montreal.

Located in one of the most dynamic regions of France, our Ubisoft studio is settled between lake and mountains, in the middle of Annecy, also known as the Venice of the Alps, 30 minutes away from Geneva International Airport. Having a studio with perpetual growth, we are focusing on our team members' commitment and team work. Our ambition relies on innovation and excellence with an ultimate purpose: our players' satisfaction!

La Forge

As Ubisoft's research and development group, La Forge brings together experts from the industry and academic sector to prototype technological innovations and improve the game-making process. With this focus on applied research, we aim to fill the gap between theory and practice, while contributing to solving real-world problems through scientific publications.

Job Description

Reinforcement Learning primarily revolves around the development of algorithms that empower agents to independently learn behaviors. The common objective is to identify an optimal policy capable of tackling particular tasks, often stipulated through a reward function. When applied to video games, these methods enable the training of effective bots for a variety of specific roles, such as navigation, marksmanship, and more.

Typically, RL models assume that the environment dynamics, the task to be accomplished, the available actions, and other essential parameters are well-defined. In our scenario, this implies waiting for the final version of the game to be available before initiating bot training, as the production process of a video game is characterized by a highly incremental nature.

To ensure the sustained efficiency of bots throughout the entire production process, it is essential for them to swiftly adapt whenever a new game version is introduced, which may entail changes in dynamics, the introduction of new abilities, and more. This requires an extension of RL techniques that enables them to incrementally adapt to evolving tasks. This challenge, commonly referred to as Continual Reinforcement Learning, remains unsolved and constitutes the central focus of the proposed internship.

Within this internship, the selected candidate will become a part of our Continual Learning Group, actively engaging in the creation of novel models capable of rapid adaptation to various tasks and changes in the gaming environment. The role will encompass the implementation and evaluation of recent techniques within a video game context.

Furthermore, the candidate will be tasked with refining these models, taking into account an offline setting in which player data is available, albeit in limited quantities, necessitating the system's ability to harness past information to address new challenges.

Qualifications
• You are a last year student of an engineering school or a university research master;
• You have solid knowledge in mathematics and computer science
• You have skills in machine learning, deep learning, or reinforcement learning, and have mobilized them using suitable Python libraries
• Your level of English allows you to work in an international team and to communicate easily with non-French speakers.

Skills and competencies show up in different forms and can be based on different experiences, that's why we strongly encourage you to apply even though you may not have all the requirements listed above.

Jobcode: Reference SBJ-g3q9k5-3-139-234-41-42 in your application.