Extropic
Research Scientist - Machine Learning
$150K–$250K
San Francisco · On-site · Full-time · Posted Jun 30, 2026
- Visa sponsorship
- Employment offer is conditioned on authorization to receive controlled items subject to U.S. export control laws and regulations, including EAR.
About the job
Develops probabilistic machine-learning theory and models, demonstrates their scaling, and deploys performant models using Extropic’s hardware.
What you'll do
- Derive theory for probabilistic models and learning rules, and scale experimentation to explore model designs.
- Implement, visualize, and evaluate architectures, training algorithms, and benchmarks; publish research and share insights with the hardware team.
- Create production models for domain experts using customer data.
What they ask for
- Experience with scientific Python and at least one deep learning framework: PyTorch, JAX, TensorFlow, or Keras.
- Strong foundations in probability and linear algebra; familiarity with deep learning theory, over-parameterization, and scaling laws.
- Publications at top ML conferences, including NeurIPS, ICML, ICLR, or CVPR.
- Experience training high-performance models and deploying models, with familiarity with the relevant infrastructure.
Nice to have
- Experience with probabilistic graphical models, energy-based models, or diffusion models.
- Experience with numerical methods in differential equation solvers, message passing or graph neural networks, information geometry, random matrix theory, or computational Bayesian methods.
Pay
$150K – $250K USD base salary
Tools
Python, PyTorch, JAX, TensorFlow, Keras, Slurm, Ray, Weights & Biases
AI focus
Research and develop probabilistic ML theory, training methods, and models, including energy-based and diffusion models; empirically assess scaling and deploy performant models.
Model-written summary · the original posting is authoritative