LabelboxUnited StatesDevelops advanced systems for human-in-the-loop AI data alignment using RLHF/DPO, improves data quality, and builds AI-assisted labeling tools. Requires Master's/PhD, 3+ years ML experience, Python/PyTorch proficiency, and top-tier publications.
Annual base salary range: $250,000—$300,000 USD (varies by skills, experience, location; excludes equity/benefits).
Researcher designing and running experiments on chain-of-thought monitorability in frontier LLMs to support scalable oversight and alignment. Requires strong empirical ML expertise with LLMs, deep interest in model behavior/alignment/interpretability, and ability to translate ambiguous questions into concrete experiments.
As a Simulation Researcher/Engineer, you will design and build simulation environments for training general-purpose robot policies. This role involves working with generative models and classical physics simulation, developing differentiable pipelines, and driving asset generation.
As a Research Scientist on the World Models team, you will invent next-generation world model architectures with a focus on controllability and physical consistency, develop controllability mechanisms, and define and own metrics for physical fidelity and action-following.
Senior researcher studies how training choices shape aligned behavior in frontier models, developing synthetic data, evaluation loops, and experiments to ensure durable, robust tendencies like honest reasoning and instruction-following.
Designs and runs experiments to improve AI model intent alignment, honesty, calibration, and robustness using RL and empirical ML methods. Trains/evaluates large models like LLMs and integrates techniques into production workflows.