researcher in self-improving agentic systems

Meta builds technologies that help people connect, find communities, and grow businesses. Its apps include Messenger, Instagram, and WhatsApp, and it is developing augmented and virtual reality experiences to advance social technology.

*Instagram и Facebook принадлежат компании Meta Platforms Inc., деятельность которой признана экстремистской и запрещена на территории РФ

• Assist in designing, implementing, and evaluating self-improving agentic systems and AI research agents
• Contribute to research on reinforcement learning and LLM post-training methods, including reward design, preference optimization, and RL from execution feedback
• Collaborate with senior researchers to explore approaches to autonomous experimentation, code generation, and optimization
• Help build and maintain evaluation harnesses and benchmarks for agentic capabilities
• Document findings and prepare research outputs for publication

• Currently pursuing a Ph.D. in Machine Learning, Artificial Intelligence, Computer Science, Information or Multimedia Retrieval, Reinforcement Learning, Mathematics, or a relevant technical field
• Experience with Python, C++, C, Java, or another related language
• Experience with deep learning frameworks such as PyTorch or TensorFlow
• Experience building systems based on machine learning and/or deep learning methods
• Будет плюсом: eagerness to learn about advanced AI concepts such as reinforcement learning, LLM post-training, agentic scaffolding, and self-improvement, excellent problem-solving and communication skills

• Contract Worker