Research Scientist Intern, Meta Recommendation Systems (PhD)

Meta is seeking Research Interns to join our research team within Meta Recommendation Systems (MRS). This team brings together a world-class team of AI Research Scientists and Engineers. The team is exploring and advancing SoTA technology, centering around AI and ML, to arrive at a unified infrastructure and model service to leapfrog recommendation systems (Recsys), its theory and algorithms across Meta (Facebook, Instagram and WhatsApp, the “Family of Apps” (FoA)). Our interns will have the opportunity to perform cutting-edge research on AI for Recsys with the potential to have an impact at Meta’s scale.

Our vision is to leverage AI frontier models and technology in all aspects of recommendation systems including but not limited to deep understanding of multi-modal content (especially images/videos, posts, user interactions), modeling user interests and preference, agentic AI for recsys, recursive self-improvement for recsys, values of the ecosystems. We contribute to the mission of connecting users to the content they enjoy, are inspired by, and that they want to see more of. To this end, we conduct cutting-edge research using the complete suite of audio, visual, text and metadata signals associated with posts and videos to improve recommendation relevance across all surfaces and provide a better, more meaningful experience to users. We build tools, create frameworks and train models that we deploy together with product and infrastructure teams to gain adoption across the FoA. We also publish scientific papers to help advance the state of the art in all aspects of the technology stack.

Our team offers twelve (12) to twenty-four (24) weeks long internships and we have various start dates throughout the year.

Responsibilities

Initiate and lead efforts towards long-term ambitious research goals, while identifying intermediate milestones in the area of recommendation systems and models, user and content understanding and multi-modal (video, audio, and text) LLM analysis for classification and relevance use cases. Conduct original research that can eventually be applied to Meta product development, engage with the wider research community, including publishing and releasing open source software where appropriate. Design, train and support AI/ML libraries and models to implement new features and functionality for use internally at Meta. Collaborate with researchers and cross-functional partners including communicating research plans, progress, and results.

Qualifications

Currently has or is in the process of obtaining a Ph.D. degree in Computer Science, Machine Learning, Computer Vision, Artificial Intelligence, or relevant technical field Must obtain work authorization in the country of employment at the time of hire and maintain ongoing work authorization during employment Experience with Python, with experience in machine/deep learning libraries such as Pytorch and Transformers Familiarity with AI/ML modeling and algorithmic techniques (e.g., various components of multimodal LLM, RAG, LSTM, GRU, Transformers, RL, and/or its acceleration for large scale use cases) and new emergent technologies such as agentic AI, RSI, autoresearch, etc Experience building systems based on machine learning and/or deep learning methods Intent to return to the degree program after the completion of the internship Proven track record of achieving significant results as demonstrated by grants, fellowships, patents, as well as publications at leading workshops or conferences such as NeurIPS, ICLR, KDD, ICML, SIGIR, WSDM, RecSys, CIKM, CVPR, ECCV/ICCV/WACV, ACL, EMNLP, ICASSP, CoLM, or similar venues Experience working and communicating cross functionally in a team environment Prior research or project experience in sequence modeling, recommendation systems, or user modeling, RL, etc Prior research or project experience about LLMs (e.g., pre/mid-training, alignment, post-training, RLVR, etc.) and agents (e.g., meta-harness, self-evolving agents, etc.) Demonstrated software engineer experience via an internship, work experience, coding competitions, or widely used contributions in open source repositories (e.g. GitHub)