Language Research Scientist

Description

We are seeking a technically skilled GenAI scientist to join our team focused on Large Language Model (LLM) agents and model post-training, with a particular emphasis on reinforcement learning (RL). This role will be close to product applications and user impact, requiring full-stack knowledge.

Responsibilities

Design, implement, and optimize LLM-based agents for a variety of applications, leveraging the latest advances in generative AI Apply reinforcement learning algorithms to improve LLM performance, safety, and alignment Integrate models and orchestrations in production Collaborate with cross-functional teams (research, engineering, product) to deploy and evaluate LLM agents in real-world scenarios Analyze and interpret experimental results, iterate on model architectures, and drive continuous improvement Contribute to the broader AI/ML community at Meta through knowledge sharing, code reviews, and technical mentorship Lead and contribute to research and development of post-training methods, including RLHF (Reinforcement Learning from Human Feedback), reward modeling, and other feedback-based approaches

Qualifications

Currently has or is in the process of obtaining a Ph.D. degree in Computer Science, Artificial Intelligence, Generative AI, or a relevant technical field Bachelor’s degree in Computer Science, Engineering, or a related technical field (or equivalent practical experience) Good programming skills in Python and familiarity with large-scale distributed training Familiarity to learn new programming languages quickly Can design, implement, and evaluate RL algorithms in production or research settings Problem-solving, communication, and collaboration skills Experience with RLHF, reward modeling, or other LLM post-training techniques Experience working in cross-functional teams Track record of publications or contributions to open-source projects in LLMs, RL, or related areas Familiarity with safety, alignment, and evaluation challenges in generative AI