We are doing research on deep learning research related to memory that would be suitable to extend the knowledge base of neural network for personalization, lifelong learning and handling large contexts. The team covers a broad range of topics, including LLM, text-time training, memory controller, indexing structures.
Responsibilities
Conduct research to advance the state of the art in architecture and memory related topics Consistently and sustainably advance the state of the art for your problem, including setting and executing against roadmaps for 6-month plus timeframes Collaborate with different cross-functional teams across the globe in research and product
Qualifications
PhD in Computer Science or a related field with published projects in the fields of machine learning, deep learning, robotics, large language models and/or computer vision Proven development skills in Deep Learning, working with PyTorch or TensorFlow Experience developing LLM algorithms or infrastructure in Python or C/C++ Significant contributions to impactful work such as open-source models Publications at peer-reviewed conferences, e.g. ICLR, ICML, NeurIPS