Research Scientist Intern, Integrity (Trust and Safety) AI (PhD)

Meta•Seattle, WA

7d•$7,650 - $12,667

About The Position

Meta is seeking a Research Scientist Intern to join the Integrity (Trust and Safety) team within Facebook. The team develops applied machine learning solutions to detect harm from bad actors, including enforcement decisions involving groups, creators, pages, and appeals. Its work spans per-post classifiers and agentic AI systems that reason over an entity’s content, members, administrators, and history, applying written policy to detect adversarial actors. Research problems include improving agentic AI performance for bad actor detection, distilling large reasoning models into cost-efficient solutions, and improving shared multi-harm models without regressing on other harms. Interns own a focused research problem end to end, from framing through experiments and publication, guided by research scientists and machine learning engineers. The role includes end-to-end machine learning workflows, AI-driven agentic workflows using tools such as Gemini, Claude, and Muse, and scaled solutions involving data analysis, feature engineering, model development, and deployment.

Requirements

Currently has or is in the process of obtaining a Ph.D. in Computer Science, Artificial Intelligence, Machine Learning, Natural Language Processing, Computer Vision, Statistics, or a 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.
Experience with deep learning frameworks such as PyTorch.
Machine learning research experience, including training or fine-tuning deep learning models and designing experiments to evaluate them.
Familiarity with large language models and AI-native development, such as Claude or Codex.

Nice To Haves

Intent to return to the degree program after completing the internship or co-op.
Research achievements demonstrated by publications at leading venues such as NeurIPS, ICML, ICLR, ACL, EMNLP, NAACL, COLM, CoLLAs, KDD, WWW, or ICWSM.
Research or project experience in one or more of continual or lifelong learning, multi-task learning and model merging, active learning, knowledge distillation, LLM post-training (SFT, preference optimization, RL), uncertainty estimation and calibration, learning from AI feedback or LLM judges, multimodal modeling, or sequence/graph modeling of user behavior.
Experience applying machine learning to trust and safety, content moderation, fraud, or abuse detection.
Experience with large-scale or parameter-efficient training, such as LoRA or distributed training.
Demonstrated software development experience through internships, work experience, coding competitions, or widely used open-source contributions.
Experience working and communicating cross-functionally in a fast-paced environment, framing real-world problems as rigorous machine learning problems, and driving iterations from clear analysis.

Responsibilities

Perform research to advance trustworthy, efficient AI systems for integrity and safety.
Develop methods for training, distilling, and continually updating large language and multimodal models that make policy decisions about entities and actors.
Design rigorous experiments and evaluations using expert human labels, LLM judges, and offline replays, including methods that remain valid when labels are scarce or expensive.
Analyze and improve the efficiency, calibration, robustness, and reliability of models used in production enforcement.
Collaborate with researchers, engineers, and policy partners, communicating research plans, progress, and results.
Publish research results and contribute to research applicable to Meta product development.
Own a focused research problem end to end, from framing through experiments to publication.
Work on end-to-end machine learning workflows, including data analysis, feature engineering, model development, and deployment.